The geographic distribution of species is the typical metric for identifying priority areas for conservation. Since most biodiversity remains poorly studied, a subset of charismatic species, such as primates, often stand as surrogates for total biodiversity. A central question is therefore, how effectively do primates predict the pooled species richness of other mammalian taxa? We used lemurs as indicator species to predict total non-primate mammal community richness in the forest ecosystems of Madagascar. We combine environmental and species occurrence data to ascertain the extent to which primate diversity can predict (1) non-primate mammal α-diversity (species richness), (2) non-primate complementarity, and (3) non-primate β-diversity (species turnover). Our results indicate that primates are effective predictors of non-primate mammal community diversity in the forest ecosystems of Madagascar after controlling for habitat. When individual orders of mammals are considered, lemurs effectively predict the species richness of carnivorans and rodents (but not afrosoricids), complementarity of rodents (but not carnivorans or afrosoricids), and all individual components of β-diversity. We conclude that lemurs effectively predict total non-primate community richness. However, surrogate species alone cannot achieve complete representation of biodiversity.