Abstract Reproductive capacity of small mammals is affected by many factors, such as climate, habitat and social stress, etc. Population density is one of the key factors that affect the reproductive capacity of female rodents. However, little has been reported on the effect of breeding under existing different population densities and different durations. Here, we compared the reproduction of adult Microtus fortis at different densities for long or short durations. We selected 144 adult voles, which were randomly divided into 4 groups (sex ratio 1:1): 2 voles per cage and housing for 90 days [group LL (low density, long time)], 8 voles per cage and housing for 90 days [group HL (high density, long time)], 8 voles per cage and housing for 20 days [group HM (high density, middle time)], 8 voles per cage and housing 10 day [group HS (high density, short time)]. Then one male and one female animal from the same group were randomly paired. The litter size, pregnancy rate, pregnancy frequency and farrowing interval were recorded every day until 180d. The experiment data were analyzed in 3 periods: the first stage was 0-90d (the treatments under different densities and different housing time), the next stage was 90-109d, (The transition period)，and the last stage was from 109d on (normal pair breeding period). One-way analysis of variance (One-way ANOVA) was used to analyze the mean litter size, and a Chi square test (χ2) was used to compare the frequency data. Results showed that the mean litter size decreased significantly after a long period of high density treatment. No significant differences have been found among groups in mean pregnant frequency. The pregnancy rates significantly increased from group HL to group HM and then to group LL and HS. In the first stage, breeding was found only in group LL. The pregnancy rates in the transition period were significantly different among the groups, the values of group LL and group HS were significantly higher than group HL. The mean litter size of group HS was the highest, that of group HL was the lowest, and there were no significant differences among groups. The pregnancy rates and mean litter size in the last stage were significantly different among the groups, and the values of group LL and HS were always higher than group HL and HM. No significant differences had been found among groups in the mean pregnant frequency, mean litter size and the farrowing interval. Taken together, there were great differences in density-dependent effect for different durations. For Microtus fortis, reproduction was more inhibited by a 20d-high-density environment than a 10d-high-density environment. After the recovery of low density, the hysteresis effect of population density constraints exists.