We reversed the experimental treatments of a previous study on food addition in the tropical mouse, Peromyscus mexicanus; food addition was suspended on three grids on which animals had been supplemented for 1 year (postfood treatment), and food was added to two previously unsupplemented grids (food treatment). One control grid remained unmanipulated (control treatment). We assessed the effect of a change in food treatment on the reproductive characteristics of adult females and on the maturation of juveniles among animals that were present in both studies. We found little evidence of residual treatment effects from the previous study. Instead, the frequency of litter failure among females in the postfood group increased to the level of control females, while the failure rate decreased among females in the food group. Furthermore, regardless of conditions in the previous study, nulliparous females provided with a food supplement in the present study were significantly more likely to produce a first litter (33%) than nulliparous females in either the postfood or control groups (15 and 0%, respectively). However, the intermediate response of females in the postfood groups suggests a possible residual effect from the previous study. The pregnancy rate and the size of successful litters (at least one young weaned) were unaffected by the food treatment.