Estrogen has been shown to play a role in modulating social recognition memory. However, the literature regarding the influence of estrogen on social memory is sparse and only covers two experimental manipulations: acute injections and receptor knockout. Long-term effects of estrogen replacement on social investigation and social recognition are unknown. Furthermore, existing social recognition protocols focus on memory of very short durations (<2 h). In the present study, we examined long-term effects of estrogen replacement on both short- (<30 min) and long-term (24 h) social recognition in ovariectomized female C57BL/6 mice by implanting 60-day time-release pellets containing physiological doses of estradiol (0, 0.18, or 0.72 mg of 17beta-estradiol). After 55 days of treatment, evidence of social recognition memory, measured by 24-h habituation, was found only in mice receiving the 0.72-mg pellet. This result is remarkable as previous reports indicate that individually-housed untreated rats and mice do not show habituation beyond 2 h. Our study further revealed that estrogen also increased frequencies of baseline social investigation without affecting general activity levels and decreased delayed post-swim-stress serum corticosterone concentration. Thus, these results suggest that long-term estrogen replacement increased the interest in social interaction as well as decreased stress responses. It is likely that the 24-h habituation observed in the estrogen replacement group is mediated jointly by the non-mnemonic effects of estrogen on the behavior displayed during the stage of memory encoding as well as mnemonic effects during the stage of memory consolidation.