Retrospective epidemiologic evaluation of 27 years of colony data (1966–1993) from a large breeding center for ruffed lemurs (Varecia variegata) was used to quantitatively assess the risk imposed by different management variables on reproductive success (littering) and infant viability (litter death). Logistic regression analysis was used, both with and without a term to account for extra-binomial variation in the data (random effects) associated with repeated assessments of individual lemur dams during the study period. Population-based littering rates are defined and shown not to fluctuate substantially over time, although the proportion of females given the opportunity to breed was drastically reduced in 1986 due to changes in the species management strategy and limited breeding center space resources. No management-related factors were found to be associated with littering success. However, young age of dam (odds ratio = 4.3, P = 0.016) and housing in natural habitat enclosures (odds ratio = 6.5, P = 0.003) significantly predicted complete litter mortality by 48 hr post-partum. Varecia females >15 years of age (8 dams, 42 observation-years) produced 18 litters and 37 infants, 31 of which survived at least 48 hr. Use of established epidemiologic methods to analyze data from existing colony records should be used to help direct the management of this endangered species toward maximally efficient goals, especially in light of current conservation resource limitations. © 1994 Wiley-Liss, Inc.