IFS Therapy: How Internal Family Systems Works and the Evidence

IFS therapy, short for Internal Family Systems therapy, is a talk therapy for mental illness and related distress that treats the mind as a system of sub-personalities, or "parts," distinct from a core Self^[2]. Its developer, Richard Schwartz, was a structural and strategic family therapist who began applying family systems ideas to his clients' inner experience about four decades before a 2023 history of the model in Family Process^[1].

The approach has spread quickly: more than 12,000 clinicians had been trained in IFS by 2022, according to a 2025 study in Frontiers in Psychiatry^[2]. Its published trials include studies in PTSD, depression, substance use and rheumatoid arthritis^[3][5][2][4].

IFS describes itself as non-pathologizing. It views the mind as an ecology of "no bad parts," each with its own motivations, sensations, thoughts and emotions^[2].

Schwartz has described the model as fostering clients' acceptance of all parts of themselves "no matter how destructive," on the premise that this acceptance is what allows those parts to change^[7]. In IFS the focus is not directly on the specifics of a trauma narrative, on symptom management or on reducing substance use^[2].

The IFS model loosely sorts parts into three categories by their therapeutic role: managers, exiles and firefighters^[8].

• Managers try to adapt the person to the demands of the outside world, often through patterns that rationally structure everyday life^[8].
• Exiles are parts burdened with severe negative emotion from past traumatic experiences^[8].
• Firefighters, like managers, try to keep exiles out of conscious awareness, for example by distracting the person through overeating or overwork^[8].

The PARTS group program, tested in a randomized trial reported in 2026, framed defense mechanisms as "protectors" of the internal system and the parts carrying the burden of past trauma as "exiles"^[3]. It described the Self as having "curious, clear and compassionate" qualities^[3].

Parts language has also been studied outside therapy. In a mental-training study of healthy adults aged 20 to 55, participants identified an average of 11 distinct inner parts over a 3-month module, and 74% of the parts could be clearly classified into established IFS-derived types^[8].

IFS relies on non-directive, inquiry-based methods: internal narrative dialogue, contemplative practice and visual imagery^[2]. For trauma, it includes participant-titrated imaginal exposure, meaning the client sets the pace of contact with painful material^[2].

The 16-week PARTS curriculum shows the typical sequence^[3]. Participants learned the framework, practiced accessing the Self, learned to recognize parts and "un-blend" from them, worked with protectors, then learned to access and help exiles before planning for continued healing^[3].

Formats in trials vary. The rheumatoid arthritis trial combined group meetings of 8 to 10 patients with 15 biweekly 50-minute individual sessions over 36 weeks^[4].

The PARTS trial delivered 16 weekly online group sessions plus 8 individual sessions^[3]. A 2025 pilot for PTSD with substance use ran 12 weekly groups with 6 individual counseling sessions^[2].

IFS has preliminary evidence, not established efficacy. Its trials are small, and the randomized PTSD trial with an active control found no advantage for IFS over the comparison program^[3].

In that trial, 60 adults with PTSD were randomized to online group IFS (PARTS) or Nature-Based Stress Reduction for Trauma Survivors, a program built around watching and discussing nature videos^[3]. The investigators hypothesized IFS would be superior, but both arms showed significant PTSD symptom reduction without significant group differences, according to a 2026 secondary analysis in Scientific Reports^[3].

For depression, a pilot trial randomized 37 college women to IFS or treatment as usual, either cognitive behavioral therapy (CBT) or interpersonal therapy^[5]. Depressive symptoms declined in both groups with no significant difference in the size or rate of change, which the authors called preliminary evidence of efficacy^[5].

The rheumatoid arthritis trial randomized 79 patients to 9 months of IFS or to an education control that received mailed materials^[4]. The IFS group reported less overall pain and better physical function after treatment, and one year later still showed gains in self-assessed joint pain, self-compassion and depressive symptoms^[4].

That trial found no sustained improvement in anxiety, self-efficacy or disease activity, which rheumatologists blinded to group assignment rated^[4].

Uncontrolled pilots report larger changes. An earlier 15-person PARTS feasibility study reported PTSD effect sizes of d = −0.7 at week 16 and d = −0.9 at week 24^[2].

In a 10-person pilot of adults with PTSD and a substance use disorder, scores on the PTSD Checklist for DSM-5 (PCL-5) fell 1.7 points per week and 70% of participants were retained at 12 weeks^[2]. Some conditions have no IFS outcome data at all: a 2026 systematic review of 12 studies on borderline personality disorder with depression or anxiety found none that tested IFS^[9].

IFS is not among the treatments in the American Psychological Association's PTSD guideline. Created in 2017 and last updated online in April 2025, it recommends three first-line therapies, all forms of CBT: general CBT, cognitive processing therapy and prolonged exposure^[6].

The guideline conditionally recommends three more psychotherapies (cognitive therapy, EMDR and narrative exposure therapy) and four medications^[6]. This site's reviews of EMDR for post-traumatic stress and CBT for depression summarize their trial records.

Direct comparisons are scarce. The only IFS trial found for this article that used CBT or interpersonal therapy as the comparison enrolled 37 women and found no difference^[5].

No trial identified for this article compared IFS head to head with cognitive processing therapy, prolonged exposure or EMDR for PTSD. The broader menu of options is covered in PTSD treatment options, and another experience-focused approach is reviewed in somatic therapy.

Small samples. The randomized trials of IFS enrolled 37, 60 and 79 participants^[5][3][4]. Samples that size can miss modest differences between treatments.

Developer involvement. Schwartz is a co-author of the rheumatoid arthritis trial^[4]. The PARTS trial was funded by the Foundation for Self Leadership, which also awarded a Richard C. Schwartz Research Fellowship to the secondary analysis's lead author; the authors state the funders had no role in that analysis^[3].

Attention effects. In the arthritis trial, the IFS group received 15 individual sessions and 12 group meetings while controls received a 1-hour meeting and mailed materials^[4]. The authors acknowledged that contact and attention alone could account for the benefit^[4].

No edge over an active control. Against a matched nature-video program, IFS performed about the same, and a secondary analysis found that a reduced sense of isolation was associated with symptom improvement in both groups^[3]. The authors suggest that this shared group element may account for part of the benefit in both arms^[3].

Gaps in the evidence. As of 2025 there were no efficacy studies of IFS for substance use disorder or of group IFS for co-occurring PTSD and substance use^[2]. Proponents cite high acceptability, such as an 86% acceptability rating in one 10-person pilot^[2], but acceptability is a different question from efficacy.