We studied testes size in free-living and laboratory-born commensal and non-commensal populations of various Mus species (M. musculus musculus, M. m. domesticus, M. spicilegus, M. spretus, M. macedonicus, and laboratory mice). We found no apparent differences between wild-caught and laboratory-born individuals, or among commensal, non-commensal, and laboratory populations of M. musculus. There were, however, considerable differences among the species studied. The highest values of relative testes size were found in the aboriginal species M. spicilegus (4.4% and 2.9% for wild and laboratory populations, respectively), followed by those of M. macedonicus (from 1.7% to 0.9% for various samples) and M. spretus (1.5%). All thirteen samples representing various populations of Mus musculus exhibited smaller testes (0.7–1.0%), and finally the three lowest mean values came from laboratory mice (0.5–0.7%). It is very surprising that aboriginal species, in particular M. spicilegus, which is widely considered to be monogamous, have relatively larger testes than the polygynous/promiscuous M. musculus. This result is in apparent contradiction to the current views on evolutionary forces affecting testes size, and suggests that there could be another uncontrolled factor obscuring the relationship between testes size and multiple paternity. This raises a question concerning the proper interpretation of social organisation in the genus Mus.