Music Therapy: How It Works and What the Evidence Shows

Music therapy is the clinical use of music, by a credentialed professional within a therapeutic relationship, to help people reach individual health goals, such as improving communication, movement, mood, or coping. Sessions can involve singing, playing instruments, improvising, writing songs, moving to rhythm, or listening and discussing music, and no musical skill is needed. Music therapists work with people of all ages in hospitals, schools, mental health services, rehabilitation, dementia care and palliative care.

Research supports music therapy for some outcomes, such as walking in Parkinson's disease and stroke rehabilitation, short-term improvement in depressive symptoms, and anxiety and pain in medical settings, though many studies are small and effects vary. This guide covers the history, neuroscience, main approaches, techniques, applications and evidence, along with the limits of what music therapy can do.

Understanding Music Therapy

Music therapy is a healthcare profession that uses music and its elements (rhythm, melody, harmony and timbre) to address physical, emotional, cognitive and social needs. Unlike listening to music for enjoyment, it involves a trained professional using music systematically, based on assessment, to work toward specific goals. Musical abilities are sometimes preserved when other abilities are damaged, for example singing in some people with aphasia, and practice can support neuroplasticity, which is one reason music can reach people who find verbal approaches difficult.

The relationship through music

The therapist uses musical skill not to perform for the client but to interact with them. With a non-verbal child, that might mean matching their vocal sounds and movements with improvised music to create a dialogue without words. With an adult, it might mean writing a song together about a difficult experience; with a person with dementia, singing familiar songs together. The therapist's ability to match and respond to a client's state through music, for example a steady rhythm for someone who is agitated, can help the person feel heard and more settled.

A bodily and cultural experience

Music is felt as vibration and movement as well as heard, so deaf people can take part in rhythm-based work through vibration. Breathing and movement tend to follow tempo, and therapists use this: slow, predictable music to support relaxation, a steady beat to organize movement in Parkinson's disease, familiar melodies to prompt memories in dementia care.

Music is also deeply cultural. A drum circle may feel meaningful to one person and alien to another, and religious music may comfort one client and exclude the next. Therapists aim to learn each client's musical culture and preferences rather than imposing their own.

Where it is practised

Music therapy takes place from neonatal intensive care units to hospices, and in psychiatric wards, schools, rehabilitation centers, clients' homes and community settings. Sessions can be individual or group, structured or improvisational, active or receptive. Practice is informed by research: assessment tools, protocols for specific conditions and outcome measures sit alongside the artistry of the work.

History and Evolution of Music Therapy

Ancient and traditional uses

Music has been linked with healing since ancient times. Sources from Egypt and Mesopotamia describe chanting in healing rituals, the Hebrew Bible describes David playing the lyre to soothe King Saul, and Plato and Aristotle wrote about music's influence on emotion and character; later traditions credited Pythagoras with using melodies to calm or rouse the emotions, though these accounts are partly legendary. Many cultures use drumming and chanting in healing ceremonies, traditional Chinese medicine linked five tones to organs and emotions, and Indian classical music associates ragas with times of day and moods. Research confirms that music affects arousal, attention and mood, but the specific beliefs of these traditions are a separate matter and mostly untested.

Early Western medicine

Robert Burton's The Anatomy of Melancholy (1621) discussed music's effects on mood. The earliest known US article on the topic, "Music Physically Considered," appeared in Columbian Magazine in 1789, and medical students Edwin Atlee (1804) and Samuel Mathews (1806) wrote early American dissertations on music's effects on mind and body. In the early 20th century, Eva Augusta Vescelius founded the National Society of Musical Therapeutics (1903), Margaret Anderton taught early college courses on music in healthcare at Columbia University (1919), Isa Maud Ilsen founded the National Association for Music in Hospitals (1926) and Harriet Ayer Seymour founded the National Foundation of Music Therapy (1941).

The World Wars and professionalization

During and after the World Wars, musicians performed in military and veterans' hospitals, and staff observed benefits for soldiers with physical and psychological injuries, including "shell shock," a term covering conditions now understood to include PTSD. The need for people trained in both music and clinical work became clear. Michigan State University created the first undergraduate music therapy degree in 1944, and E. Thayer Gaston, often called the "father of music therapy," established a graduate program at the University of Kansas in 1946, emphasizing research and the principle that music is a human behavior shaped by culture.

The National Association for Music Therapy (NAMT) formed in 1950 and the American Association for Music Therapy in 1971. The Certification Board for Music Therapists followed in 1983, and in 1998 the two associations merged into the American Music Therapy Association (AMTA). Meanwhile, distinct approaches developed: Paul Nordoff and Clive Robbins' improvisational Creative Music Therapy, Helen Bonny's Guided Imagery and Music, and Mary Priestley's Analytical Music Therapy.

Neuroscience and global growth

From the 1990s, Michael Thaut and colleagues developed Neurologic Music Therapy, grounding techniques such as rhythm-based gait training in research on how movement synchronizes with rhythm. Brain-imaging and neurochemistry research, such as studies of dopamine release during pleasurable music, offered possible explanations for music's effects on mood and motivation, though findings on other hormones are mixed.

The World Federation of Music Therapy was founded in 1985, and music therapy is now practised in many countries, often adapted to local traditions, as in community music therapy in South America. Recent developments include telehealth, accelerated by the COVID-19 pandemic, digital and adaptive instruments, and ongoing debates about insurance coverage, workforce diversity and keeping humanistic values in medicalized systems.

The Science of Music and the Brain

A whole-brain activity

Listening to and making music engages auditory areas and also motor systems, emotion and reward circuits, memory networks and frontal regions involved in attention and planning. Music processing is spread across both hemispheres rather than concentrated on one side as much language is, and different elements such as melody, rhythm and lyrics rely partly on different networks. This distributed processing helps explain why brain injury rarely removes all musical abilities, and why some people with aphasia can sing words they cannot speak. Musicians show structural brain differences linked to training, but whether short-term therapeutic music-making produces comparable changes is less clear.

Rhythm and entrainment

Entrainment is the tendency of movement and other bodily rhythms to synchronize with an external beat. Auditory and motor regions are closely connected, and motor areas activate even when people only listen to rhythm. Because a beat is predictable, it gives the brain an external timekeeper for planning movement. This is the basis of Rhythmic Auditory Stimulation: in Parkinson's disease, where internal timing is impaired, and after stroke, walking to a steady beat can improve speed, stride length and symmetry.

Emotion, reward and memory

Music engages emotion and reward systems. A 2011 study by Valorie Salimpoor and colleagues found dopamine release in reward areas both when people anticipated and when they experienced peak emotional moments in music they loved, which may help explain music's motivating effects. Claims about other chemicals, such as oxytocin or endorphins during group singing, rest on smaller studies with mixed results.

Musical memory is often well preserved. People with dementia may still sing songs or play instruments learned decades earlier, perhaps because these draw on procedural and emotional memory as well as the systems most affected by the disease. People also tend to remember music from adolescence and early adulthood especially well (part of the "reminiscence bump"), so music from those years is often used in reminiscence work with older adults.

Attention and stress

Following or playing music requires attending to several streams at once, and its predictable structure can help organize attention, which is one rationale for attention-training exercises in rehabilitation. Music also influences the autonomic nervous system: slower, predictable music tends to support relaxation, while fast, complex music increases arousal. Studies in medical settings have linked music to lower anxiety and in some cases lower cortisol, and coordinating breathing with musical phrasing is used to support calmer breathing, though physiological results vary between studies.

Major Approaches in Music Therapy

Nordoff-Robbins (Creative Music Therapy)

Developed in the 1960s by composer Paul Nordoff and special educator Clive Robbins, this approach holds that everyone has an innate musicality, the "music child," that can be reached even when other channels of communication are blocked. The therapist improvises, often at the piano, matching and supporting the client's sounds and movements while inviting interaction. Change is understood to happen through the musical relationship itself, and progress is assessed through recordings, for example greater rhythmic organization or longer engagement, which allows assessment of clients who cannot take part in verbal tests.

Bonny Method of Guided Imagery and Music (GIM)

Helen Bonny developed GIM in the 1970s. After a conversation and a relaxation induction, the client listens to a carefully sequenced program of mainly classical music, typically 20 to 45 minutes long, and describes the images, feelings and sensations that arise while the therapist offers minimal guidance. Afterwards the experience is discussed or drawn. Because it can evoke intense material, GIM requires extensive postgraduate training, including personal sessions.

Analytical Music Therapy

Mary Priestley developed this psychodynamic approach in the 1970s. Client and therapist alternate between talking and improvising together, sometimes on a theme drawn from the conversation, such as "play your anger." The music, including a client's choice of instrument or habit of playing barely audibly, is explored for what it reveals, and the therapist's own musical reactions are treated as information about the relationship.

Behavioral music therapy

Developed by Clifford Madsen and others, this approach applies learning principles. Preferred music can reinforce target behaviors, songs can structure routines (a song with a verse for each step of getting dressed, for example), and particular tunes can become cues, such as a cleanup song. It is often used with people with developmental disabilities.

Community music therapy

Community music therapy addresses the social and cultural side of health. Instead of the therapist as expert, community members' own musical traditions guide the work, and projects such as choirs, songwriting workshops or public performances aim to build connection, give voice to marginalized experiences and challenge stigma around mental illness or disability.

Medical music therapy

In hospitals, music therapy is used to support patients through procedures, manage anxiety and pain, and improve quality of life. Live music lets the therapist adjust tempo and volume to the patient's responses. In neonatal care, therapists may teach parents to sing lullabies matched to the infant's breathing to support calm and bonding. Neurologic Music Therapy, the most research-driven model, has its own section below.

Music Therapy Techniques and Interventions

Techniques range from strict protocols to free improvisation, and from making music to listening. Common methods include:

Some practitioners also use vibroacoustic therapy (low-frequency vibration delivered through a bed or chair) and sustained vocal toning for relaxation; claims about these, such as improved circulation or stimulation of the vagus nerve, are not well established.

Receptive and active techniques

Receptive methods involve listening to live or recorded music chosen for a therapeutic purpose, followed by relaxation, imagery or discussion. They suit people with limited energy or mobility, those who feel self-conscious, or medical settings. Active methods involve singing, playing, improvising or composing and are useful for building confidence, motor skills and social interaction. Many sessions combine both, for example starting with listening to build rapport, moving to music-making and ending with listening for closure.

Individual and group formats

Individual sessions allow close tailoring and privacy and may be the only option for people who are medically fragile. Groups use music's social nature: drum circles, choirs, ensembles and group songwriting offer peer support, turn-taking and listening practice, and are cost-effective.

Adapting techniques

The same principle can be adapted widely. Rhythmic entrainment might mean rocking an infant to a lullaby, marching for a preschooler or gait training for an adult. Music is chosen to fit the person's culture, and drumming that feels joyful to one person may have unwelcome associations for another. Adaptive instruments, switch-operated devices, visual cues and vibration allow people with physical, hearing or other disabilities to take part.

Clinical Applications of Music Therapy

This section describes how music therapy is used; the research section summarizes how strong the evidence is for each area.

Autism and developmental disabilities

Music therapy is widely used with autistic children and adults (see autism spectrum), with mixed research results. The predictability of music can be reassuring, and shared music-making offers social contact with less direct pressure. Playing together involves joint attention and turn-taking, some children can sing words they do not yet say, and functional phrases may be set to simple melodies with the musical support gradually faded. For people with intellectual disabilities, musical tasks can practise memory, sequencing and attention, with immediate feedback and a chance of success.

Mental health

For depression, active music-making can counter withdrawal and passivity, and group work reduces isolation. For anxiety, slower, predictable music and breathing paced with musical phrases are used to support relaxation, and a live therapist can adjust the music in real time. For people affected by trauma, the bounded structure of a song can give containment for difficult feelings, drumming can release physical tension, and songwriting can put experiences into words; music therapy is generally an adjunct to established trauma treatments. For people with schizophrenia, structured music-making can be grounding and social without heavy verbal demands, and reviews suggest it may help negative symptoms and social functioning when added to standard care.

Neurological rehabilitation

After stroke, rhythmic cueing supports walking, instrument playing exercises arm movements, and Melodic Intonation Therapy draws on preserved singing ability to support speech in aphasia. In Parkinson's disease, an external beat helps compensate for impaired internal timing, and singing is used to work on voice volume and breath support; groups also offer social contact. After traumatic brain injury, music is used for attention training, memory and coping with grief and changed identity. In multiple sclerosis, rhythmic work on walking and relaxation is adapted to fatigue and fluctuating symptoms.

Children and adolescents in healthcare

In pediatric hospitals, live music adapted to a child's responses is used to reduce distress during procedures, and therapists may help parents sing to their child. In neonatal intensive care, carefully calibrated music, often parents' own singing, is used to support premature infants' calm and feeding; the Pacifier-Activated Lullaby system plays a lullaby when the infant sucks, reinforcing feeding. Adolescents, for whom music is often central to identity, may engage through playlists, lyric discussion, songwriting or band work more readily than through conventional talk therapy.

Older adults, dementia and end of life

For people with dementia, familiar music from early adulthood can prompt memories, engagement and conversation. Singing may lift mood, music during personal care can make it less distressing for some people, and groups offer social contact at an appropriate level. Caregivers and family members can learn to use music to connect. In end-of-life care, music therapy is used to ease anxiety and pain perception, create legacy songs or recordings, and give families a way to be present.

Pain and medical conditions

For chronic pain, music may help through distraction, relaxation, emotional expression and a sense of control. In cancer care it is used during chemotherapy and radiation to ease anxiety and help patients cope, and to address existential concerns. In cardiac rehabilitation, music can pace exercise and support relaxation, and in respiratory conditions such as COPD, singing and wind instruments are used to practise breath control.

Neurologic Music Therapy

Neurologic Music Therapy (NMT), developed by Michael Thaut and colleagues at Colorado State University in the 1990s, applies music to non-musical brain functions. Rather than relying mainly on music's emotional or social qualities, it uses specific mechanisms: rhythmic entrainment for movement, melody and rhythm for speech, and musical structure for attention and memory. The model includes about 20 standardized techniques in three domains, each with defined protocols, which makes the approach easier to research and to fit into rehabilitation teams.

Sensorimotor techniques

Rhythmic Auditory Stimulation (RAS), the most researched technique, uses a beat as an external timekeeper for walking. Patterned Sensory Enhancement uses musical features to cue movements that are not naturally rhythmic, such as rising pitch for reaching upward or a crescendo for increasing force. Therapeutic Instrumental Music Performance positions instruments so that playing them exercises functional movements such as reaching or gripping, with the sound giving immediate feedback.

Speech and language techniques

Melodic Intonation Therapy (MIT) is used for severe non-fluent aphasia. Short phrases are intoned on two pitches with rhythmic hand tapping, and the melodic support is gradually withdrawn toward normal speech; imaging studies have reported increased right-hemisphere involvement after treatment. Musical Speech Stimulation builds on the fact that some people with aphasia can sing familiar lyrics, moving from completing song lines toward spontaneous speech. Rhythmic Speech Cuing uses a beat to pace speech in dysarthria, apraxia of speech or stuttering.

Cognitive techniques

Musical Attention Control Training practises sustained, selective, divided and alternating attention through musical tasks. Musical Executive Function Training uses improvisation and composition to exercise planning, decision-making and flexibility. Musical Mnemonics Training sets information, such as a daily routine, to melody and rhythm, because sung material is often easier to remember.

Assessment and practice

NMT uses Thaut's Transformational Design Model: assess a functional problem, set a non-musical goal, translate it into a musical exercise, then transfer the gains into everyday activities. Outcomes are measured with standard tools such as timed walking or speech intelligibility, and goals are specific and measurable in the style of medical rehabilitation. Practitioners train through the NMT Academy and work alongside physiotherapists and speech-language pathologists, complementing rather than replacing their work.

Evidence is strongest for RAS in gait rehabilitation, particularly in Parkinson's disease and stroke, and MIT has support in aphasia. Evidence for the cognitive techniques is more limited and comes mainly from small studies.

Research and Evidence Base

Neuroscience research

Brain imaging shows that music engages widespread networks involved in hearing, movement, emotion, memory and reward. Studies of melodic intonation therapy for aphasia have reported increased activity in right-hemisphere language-related areas after treatment, and research on rhythm shows that the motor system readily synchronizes with an external beat. Research on brain chemistry, such as Salimpoor's 2011 dopamine study, is suggestive, while studies of group singing and cortisol or oxytocin levels have produced mixed results. Most imaging studies are small, and these findings describe how music affects the brain in general; they do not by themselves show that music therapy treats particular conditions.

Clinical efficacy research

• Autism: an early Cochrane review (Geretsegger and colleagues, 2014; 10 studies, 165 participants) found music therapy may help social interaction and communication. A large international randomized trial published in JAMA in 2017 (the TIME-A trial, 364 children) found no significant difference in autism symptom severity between improvisational music therapy and enhanced standard care. The 2022 Cochrane update concluded that music therapy probably increases the chance of global improvement, with less certain effects on specific symptoms.
• Depression: a 2017 Cochrane review (Aalbers and colleagues; 9 studies, about 400 participants) found that music therapy added to usual care probably reduces depressive symptoms in the short term, but longer-term effects and comparisons with other psychological therapies remain uncertain.
• Pain and anxiety in medical settings: meta-analyses, including a 2016 review of 97 studies by Jin Hyung Lee, found that music interventions reduce self-reported pain and anxiety. Many of these studies tested recorded music listening chosen by medical staff (sometimes called "music medicine") rather than music therapy delivered by a credentialed therapist.
• Dementia: a 2018 Cochrane review (van der Steen and colleagues) found music-based interventions probably reduce depressive symptoms and may improve overall behavioral problems in people living in care homes, with little evidence of effects on agitation or cognition.
• Schizophrenia: a Cochrane review (Geretsegger and colleagues, 2017) found music therapy added to standard care may improve global state, negative symptoms and social functioning, especially when sessions are numerous enough.
• Stroke and Parkinson's disease: rhythmic auditory stimulation has trial evidence for improving walking speed and stride length in Parkinson's disease and after stroke.

Comparative effectiveness

Direct comparisons between music therapy and other treatments are relatively rare. Some rehabilitation trials suggest that adding rhythmic cueing to gait training improves walking more than gait training alone. Music listening is a low-cost, low-risk way to reduce procedural anxiety, but it is used alongside, not instead of, standard medical care. In dementia care, music-based interventions perform similarly to other well-run psychosocial activities, and no single non-drug approach clearly outperforms the others.

Qualitative research and mechanisms

Qualitative studies describe themes such as connection, self-expression, identity and having a "voice" when words are difficult, and ethnographic work on community music therapy shows how cultural context shapes what music means. Proposed mechanisms include rhythmic entrainment for movement, melodic structure for speech, emotional expression and regulation, attention and distraction for pain and anxiety, and social connection in groups. The therapeutic relationship and personal preference also matter: music chosen by or meaningful to the person tends to be more engaging. Separating these ingredients experimentally is difficult, and many claimed mechanisms remain hypotheses.

Special populations

In neonatal intensive care, studies such as Joanne Loewy and colleagues' 2013 trial in Pediatrics reported effects on premature infants' heart rate, feeding and sleep, and on parental stress, though results across studies are mixed and long-term developmental benefits are not established. Programs for military veterans, including group drumming and songwriting, report improvements in mood and social connection, but controlled trials are few. In palliative care, reviews suggest music therapy may improve quality of life and reduce distress, with low-certainty evidence.

Limitations and future directions

Interventions, populations and outcomes vary widely, which complicates meta-analysis. Blinding is usually impossible, many trials are small, and positive results are more likely to be published. Studies often blur the line between music therapy delivered by trained therapists and simple music listening. Priorities include larger, independent trials, clearer reporting of what was delivered, longer follow-up, cost-effectiveness studies and research in more diverse cultural settings.

Education and Professional Development

Education and clinical training

In the United States, the entry qualification is a bachelor's degree (or equivalent) from an AMTA-approved program. Programs combine music (theory, performance and functional skills on piano, guitar, voice and percussion, plus improvisation and arranging), music therapy (foundations, techniques, psychology of music, assessment, documentation and ethics) and behavioral and health sciences. Students complete supervised pre-internship fieldwork (typically about 180 hours) and then an internship of at least 1,040 hours at an approved site, working as supervised entry-level therapists.

The therapist's musical skill is functional rather than performative: simplifying songs, changing keys, improvising accompaniments and knowing repertoire across genres, generations and cultures. Clinical, interpersonal and professional competencies include assessment, treatment planning, cultural responsiveness, self-awareness and working within healthcare systems.

Credentials and regulation

The Music Therapist-Board Certified (MT-BC) credential, awarded by the Certification Board for Music Therapists, requires an approved degree and internship and a national examination. MT-BCs must complete 100 continuing education credits every five years. State regulation varies: some US states license music therapists specifically, some include them under broader licenses and some have no regulation. In much of Europe, a master's degree is the usual entry qualification, and theoretical emphasis differs between countries.

Advanced training

Postgraduate training is available in the Bonny Method of GIM, Neurologic Music Therapy and Nordoff-Robbins music therapy, and in specialties such as neonatal intensive care, hospice and palliative care and trauma-informed practice. Master's and doctoral programs prepare therapists for advanced practice, research and teaching. Current education debates include the integration of neuroscience and telehealth into training, interprofessional education with health students, and widening access to the profession for underrepresented groups.

Future Directions and Conclusion

Several developments are shaping the field. Technology offers telehealth, adaptive digital instruments, apps that may extend practice between sessions and biosensors that track physiological responses, though questions remain about preserving the therapeutic relationship through a screen and about equal access. Researchers hope that objective measures, such as heart rate variability, could complement self-report, particularly for non-verbal clients, while recognizing that such measures capture only part of what music therapy offers. Community and preventive applications in schools, workplaces and community settings are growing, as is interest in music therapy where resources are limited, which raises questions about training, quality and cultural fit.

Future research priorities include clarifying mechanisms, implementation research on how programs succeed in real services, and economic evaluations. The field also continues to balance standardized protocols with flexible, responsive practice.

In summary, music therapy has grown from ancient traditions and wartime hospital volunteering into a credentialed profession with distinct models, from improvisational and psychodynamic approaches to neuroscience-based rehabilitation. The evidence is strongest for rhythm-based gait training, with reasonable support for short-term benefits in depression, reduced anxiety and pain in medical settings and some outcomes in dementia and schizophrenia, and mixed results in autism. It is best understood as a complement to other care, offering a channel for connection, expression and rehabilitation that can reach people for whom words are difficult.