We examined memory improvement with respect to the effects of gastrin-releasing peptide (GRP) in male C57BL/6J mice under conditions of experimentally induced amnesia. GRP was administered following training in a one-trial passive avoidance test. In Experiment 1, the drug scopolamine (1 or 2 mg/kg, ip) was used to induce amnesia prior to training, and GRP (32 nmol/kg, ip) or saline (control) was administered immediately after training. Results indicate that GRP at this dose improved memory only when the dosage of scopolamine was relatively low (1 mg/kg). In Experiment 2, CO2-induced amnesia was employed. Mice were placed in a chamber filled with CO2 or air (control) immediately after acquisition training. Subsequently, they were administered either saline or GRP (32 nmol/kg, ip). Significantly longer light-dark latency was observed in all mice that received GRP (both experimental and control groups). In total, our results indicate that the effect of GRP at this dose on the improvement of impaired memory is dependent on the degree of impairment. Furthermore, because CO2-induced hypoxia is known to decrease acetylcholine release in the brain, our results also suggest that GRP and its receptor may interact with the cholinergic system in the central nervous system.