Background/Question/Methods Impacts of global change on animal life histories are projected to be diverse across space, time, and traits. These impacts will also be species-specific; however, spatially and temporally comprehensive life history trait datasets are lacking for all but the most easily-observed taxa. We took an informatics-based approach to reconstructing breeding phenology and its drivers in three clades of small mammals (Peromyscus mice, Microtus and Myodes voles, Sorex shrews) that are widespread across North America. To do this, we combined individual-level reproductive trait observations from digitized museum specimens and field censuses that collectively spanned over a century. We used this novel dataset and a mixed-model approach to reconstruct species-specific breeding phenologies in different ecoregions, and to test the importance of major environmental variables as breeding cues. Results/Conclusions Despite the heterogeneous nature of our data, we successfully reconstructed breeding phenologies in these species, which are among the most dominant in small mammal communities across North America. As expected based on physiological and life history differences, we found that breeding phenologies as well as the importance of distinct environmental cues varied substantially among taxa, as well as within taxa but among regions. Importantly, for a subset of species, we linked the differential importance of breeding cues to ecosystem-specific limiting conditions (e.g., total precipitation). Our results provide new insight into small mammal reproduction and its drivers in the wild. They also highlight the critical need to develop denser, individual-level trait-bases for secretive or hard-to-monitor taxa, including the much of the remainder of global small mammal diversity.