Why Prolonged Exposure for PTSD Works Best After Treatment Ends

When You Hesitate to Begin Exposure—and the Evidence Behind That Pause

If you work with trauma, you know the moment. The client still seems too fragile to start exposure—but staying with avoidance week after week only seems to harden the pattern. "Push into the trauma too soon and they'll bolt; avoid it too long and the avoidance calcifies." That hesitation is part clinical judgment and part the clinician's own anxiety, and it's worth naming honestly.

Here's a finding that complicates the picture in a useful way: prolonged exposure (PE) does not show its clearest benefit at the moment treatment ends. PE's real effect emerges most strongly roughly three months after termination. This isn't simply "slow therapy." It means that if you evaluate the treatment by termination-point metrics alone, you will systematically undervalue it. Below, we revisit Foa and colleagues' (1991) classic randomized controlled trial (RCT)—the study that first made this delayed effect visible—and translate it into concrete decisions you can make in the room.

What Prolonged Exposure Actually Is

Prolonged exposure, developed by Edna Foa, is an evidence-based treatment for PTSD. Its core mechanism is straightforward: by systematically and repeatedly engaging with trauma memories and avoided situations, the trauma memory stops being processed as a present-day danger signal.

PE rests on two primary components:

The question clinicians new to PE ask most often is: "If we move into exposure too fast, won't the client be retraumatized?" The concern is reasonable—but the clinical data point the other way. Exposure conducted in a safe, structured, well-paced context does not reinforce avoidance; it gradually dismantles it. What matters most, it turns out, is timing—and PE's payoff lands after the work, not during it.

The Anchor Study: Foa et al. (1991)

Foa and colleagues' (1991) trial remains a landmark in comparative PTSD research. Forty-five women with assault-related PTSD were randomized to one of four conditions—prolonged exposure (PE), stress inoculation training (SIT), supportive counseling (SC), or a waitlist control—each delivered across nine biweekly 90-minute sessions.

At the post-treatment assessment, SIT produced the largest reduction in PTSD symptoms. PE trailed it. Read that snapshot alone and the obvious conclusion is "SIT beats PE."

But at the follow-up roughly 3.5 months later, the ranking reversed. PE showed the greatest symptom reduction, and its advantage over SIT was statistically significant. PE's gains kept compounding after the last session; SIT's and SC's did not.

The clinical meaning is clear. Exposure is not a treatment that improves fastest inside the session—it's one that consolidates afterward. Memory processing and fear extinction continue to unfold in the weeks and months following the exposure work itself.

Supportive counseling, meanwhile, beat the waitlist but never gained further ground over time—an early, clean demonstration that warm listening alone rarely breaks the reexperiencing–avoidance loop at the heart of PTSD.

Three Clinical Implications of the Delayed Effect

1. Don't judge outcome by termination-point data alone

A client may finish nine sessions and say, "I don't feel that much better." That is not proof the treatment failed. PE's benefit accrues over the months following termination. Resist over-interpreting the end-of-treatment assessment as the verdict.

2. Prepare the client in advance

Explaining this pattern before you begin PE measurably supports retention.

"What we work on today won't fully resolve this week. It settles in over time."

This framing protects against clients dropping out because they conclude "it isn't working." Pair it with a heads-up that exposure work is hard and that symptoms may temporarily spike—and that this is not a sign of failure but part of the processing itself.

3. Build the follow-up session into the structure

When PE ends, schedule a follow-up session about three months out. Treat this session as part of the protocol, not an optional extra. Consistent with Foa et al. (1991), it's the point at which PE's real effect becomes visible—an opportunity to confirm progress and review it together with the client.

Single-Incident vs. Complex PTSD: How Exposure Timing Changes

The Foa et al. (1991) sample had single-incident (assault) PTSD. For clients carrying complex trauma—childhood abuse, prolonged domestic violence—applying PE directly can play out differently.

Cloitre and colleagues' (2010) RCT found that in complex, childhood-abuse–related PTSD, the strongest outcomes came when emotion-regulation and interpersonal-skills training (STAIR) preceded narrative trauma exposure. With complex trauma, the timing of exposure warrants far more deliberation than it does in single-incident PTSD.

PE's Effect Isn't Finished at Termination

Foa et al. (1991) hands clinicians a durable reference point: PE's benefit consolidates over the months after termination—not during the sessions or at their close. Evaluate the treatment by termination metrics alone and you will miss its value.

Make the pattern explicit—for the client and for yourself. "What we work on today won't fully resolve this week. It settles in over time." That single sentence keeps clients in treatment and keeps clinicians from pivoting to a different approach too soon. Tracking session-by-session ratings and a scheduled follow-up in your EHR or case-tracking system makes this delayed trajectory legible—so a quiet termination week reads as expected, not as failure.