A beneficial effect of sigma (sigma) agonists was previously described on several pharmacological models of learning impairments. We examined this effect in senescence-accelerated mice (SAM), which has been developed as a murine model of aging and cognitive dysfunction. SAMP8/Ta (P8, senescence-prone substrain), 10-12 months of age, showed significant impairments in mnemonic capacities, as compared to age-matched SAMR1/Ta controls (R1, senescence-resistant substrain). Tests included open-field behavior, spontaneous alternation performances in the Y-maze, step-down passive avoidance and place learning after repetitive training in a water-maze. Pretreatment with the sigma agonists JO-1784 (igmesine) or PRE-084, at 0.1-3 mg/kg, s.c., significantly improved spontaneous alternation and passive avoidance performances in P8. JO-1784 or PRE-084, at 1 mg/kg, also improved place learning in the water-maze, and retention, in term of escape latency. The implication of sigma sites was indicated by the lack of significant effect of JO-1783, the inactive enantiomer of JO-1784, and by the ability of BMY-14802 (5 mg/kg, i.p.) to antagonize the effects on passive avoidance of JO-1784 (0.5 mg/kg) or PRE-084 (1 mg/kg). Subchronic treatments with JO-1784 (0.5 mg/kg/day) or PRE-084 (1 mg/kg/day) during 10 days, allowed a significant improvement of learning during training in the water-maze, but retention was not significantly ameliorated. These results confirmed the interest of the SAM substrains as an experimental model for senile memory impairment and showed that sigma agonists could improve the quality of learning, although they seem less effective on long-term memory retrieval upon chronic administration.