The hypothesis that mammae number correlates with the offspring number of species can be traced back to Aristotle. However, it has never been tested rigorously. Previous efforts to investigate the relationship between these traits were of limited taxonomic scope to address mammal-wide patterns and did not consider the effect of phylogeny on species data. Here we comprehensively sample mammal diversity, analyzing data for 2,301 species, to ask: Do mammae number and litter size co-evolve and, if so, how? Using phylogenetic generalized least squares regression, we show that across Mammalia, mammae number evolves with an approximately one-to-one relationship with the maximum reported litter size of a species. Mammae number predicts litter size more strongly than other species level traits (adult body mass, gestation length, diet, and seasonality of contemporary geographic distribution) which is surprising given the usual emphasis on energetic traits and body mass to explain litter-size variation. Clades can diverge from these general patterns, revealing the influence of diverse life-history strategies upon mammary evolution. We argue that mammae number is an underappreciated constraint on fecundity and that its evolution has influenced the radiation of mammals.