Fear memory could be a tremendous burden for anxiety disorders such as post-traumatic stress disorder and phobia. The formation of fear memory refers to the phenomenon wherein a neutral stimulus (conditioned stimulus, CS; e.g., a tone) that does not initially induce fear in an individual, begins to do so after it is repeatedly paired with an intrinsically aversive consequence (unconditioned stimulus, US; e.g., a shock). Recent evidences suggest that fear memories can be updated by presenting a single cue prior to extinction during the reconsolidation time window (ret+ext). However, real-life traumatic events are usually associated with multiple different cues, and sometimes more than one sensory modality (e.g. both auditory and visual cues). Here we hypothesize that complex fear memory can be disrupted by retrieval-extinction interference in reconsolidation time window. We addressed this hypothesis by using a modified ret-ext paradigm based on a three consecutive days' experiment. Skin-conductance response (SCR) served as the measurement of fear responses. Three groups of participants underwent fear conditioning by a discrimination paradigm with partial reinforcement (30%) in day1. For the fear conditioning, two colored squares and two tones were used, one square and one tone combination served as a compound conditioned stimulus (CS+) that paired with a mild shock to the wrist (unconditioned stimulus) on 38% of the trials, whereas the combination of the other square and the other tone was never paired with shock (CS−). A day later, Group T received a reminder trial of a tone (auditory part of the compound CS+), Group P received a reminder trial of a square (visual part of the compound CS+), and Group T+P received a reminder trial of the combination of tone and square (absolute same compound CS+ as day1). 10 minutes later, all three groups underwent extinction training in which the two conditioned stimuli were repeatedly presented without the unconditioned stimulus. Twenty-four hours later, all three groups underwent extinction again (re-extinction) to assess spontaneous fear recovery. Then, all three groups received 4 unsigned shocks (US), followed by a extinction session to assess reinstatement. Current results showed that the SCR were not significantly different among the three groups in fear conditioning of the first day (F(2,34) = 0.024, p = 0.98) and extinction of the second day (F(2,34) = 0.10, p = 0.91). In the third day, Group T showed increased SCR in spontaneous fear recovery (p < 0.001) and reinstatement test (p < 0.01), whereas Group P only showed increased SCR in reinstatement test (p < 0.01), but no spontaneous recovery (p > 0.05). And Group T+P did not showed increased SCR in spontaneous fear recovery (p > 0.05) or reinstatement test (p > 0.05). Our study provided evidence that the behavioral interference during reconsolidation time window (retrieval-extinction) can block the spontaneous recovery and reinstatement of fear memory. According to the results, we also demonstrated that the compound memory can be disrupted by interrupting the reconsolidation (using the stronger individual component). These findings provided insight into how compound fear memories encoded, and have clinical implications for future PTSD treatment.