The present experiments characterized the acquisition of fear-potentiated startle (FPS) and determined the sensitivity of FPS to anxiolytic compounds in DBA/1J mice. A light (30 s) conditioned stimulus (CS) and mild footshock (0.14 mA, 0.5 s) unconditioned stimulus (US) were used. First, acquisition of FPS was examined by presenting the acoustic startle probe during and after each CS-US pairing trial, allowing for a trial-by-trial measurement of experience-dependent startle plasticity. In this novel protocol, mice showed robust acquisition (larger acoustic startle reflex in the presence of the CS) of FPS after as few as eight CS-US pairings. FPS was significantly greater when the CS and US were paired explicitly (light-paired) as compared to when both the US and CS were presented randomly (unpaired), or when the CS was presented alone (no shock), indicating pairing-dependent learning of the CS. Second, the present study assessed the sensitivity of FPS in mice to anxiolytic drugs. The GABA-A receptor agonists diazepam (3 and 6 mg/kg) and chlordiazepoxide (10 mg/kg) significantly reduced the expression of FPS post-training, as did the serotonin 1A receptor partial agonist buspirone (5 and 10 mg/kg). Furthermore, all three anxiolytic drugs reduced startle responding in a cue-specific manner and without significant changes in baseline responding. These data demonstrate a novel method of studying acquisition of FPS, and support the predictive validity of the FPS model of anxiolytic drug action in mice.