We conducted comparisons for exotic mammal species introduced to New Zealand (28 successful, 4 failed), Australia (24, 17) and Britain (15, 16). Modelling of variables associated with establishment success was constrained by small sample sizes and phylogenetic dependence, so our results should be interpreted with caution. Successful species were subject to more release events, had higher climate matches between their overseas geographic range and their country of introduction, had larger overseas geographic range sizes and were more likely to have established an exotic population elsewhere than was the case for failed species. Of the mammals introduced to New Zealand, successful species also had larger areas of suitable habitat than did failed species. Our findings may guide risk assessments for the import of live mammals to reduce the rate new species establish in the wild.