Top panel, scheme illustration of the behavioral procedure: fear conditioning (Day 1), memory tests (Day 2), trauma re-contextualization by complete re-exposure to trauma-related cues (i.e. to the tone in the conditioning context vs. partial re-exposure to the tone alone, context alone, or tone and context spaced out of 2 h; Day 3), memory tests (Day 4). a Curative effect of trauma re-contextualization with a switch from PTSD-like memory on Day 3 to normal contextual memory on Day 4: complete re-exposure to trauma-related cues in CORT-injected mice abolishes the abnormal fear response to the tone and normalizes the conditioned fear to the context (Veh vs CORT: ns). b–d Partial re-exposure to trauma-related cues (tone alone, context alone or tone and context spaced out of 2 h) on Day 3 does not modify the PTSD-like memory profile in CORT-injected mice, which still display an abnormal fear response to the tone together with a low conditioned fear to the context on Day 4 (minimum significance in the tone test: repeated measures × Veh/CORT: F2,36 = 4.755; P = 0.0147; in the context test: Veh/CORT: F1,18 = 4.570; P = 0.0465). CORT-injected mice discriminate the two contexts used, as attested by their higher freezing levels in the conditioning context than in the familiar chamber (first 2 min, 2a vs. 2b; F1,17 = 7.166; P = 0.0159). The blue and gray bars symbolize the conditioning context and the familiar/safe chamber, respectively; Pre-T: Pre-Tone; Post-T: Post-Tone. Data are presented as mean ± SEM. This experiment was repeated independently once with similar results, see Table 1 in the “Animals” section of the methods for detailed sample size (n = 9–10). Statistical significance was assessed by repeated measures (three blocks) two-sided ANOVA. *P < 0.05; **P < 0.01; ***P < 0.005. #: block × condition interaction (P < 0.05); ##P < 0.01; ###P < 0.005. Source data are provided as a Source Data file.