I analyze and summarize the empirical evidence supporting alternative hypot
heses posed to explain the evolution of rodent group-living. Eight hypothes
es are considered: two rely on net fitness benefits to individuals, five re
ly on ecological and life-history constraints, and one uses elements of bot
h. I expose the logic behind each hypothesis, identify its key predictions,
examine how the available evidence on rodent socioecology supports or reje
cts its predictions, and identify some priorities for future research. I sh
ow that empirical support for most hypotheses is meager due to a lack of re
levant studies. Also, empirical support for a particular hypothesis, when i
t exists, comes from studies of the same species used to formulate the orig
inal hypothesis. Two exceptions are the hypothesis that individual rodents
live in groups to reduce their predation risk and the hypothesis that group
-living was adopted by individuals to reduce their cost of thermoregulation
. Finally, most hypotheses have been examined without regard to competing h
ypotheses and often in a restricted taxonomic context. This is clearly an u
nfortunate situation given that most competing hypotheses are not mutually
exclusive. I suggest that in the future comparative approaches should be us
ed. These studies should examine simultaneously the relevance of different
benefits and constraints hypothesized to explain the evolution of rodent so
ciality.