Background. Frogs are generalist predators of a wide range of typically small prey items. But descriptions of dietary items regularly include other anurans, such that frogs are considered to be among the most important of anuran predators. However, the only existing hypothesis for the inclusion of anurans in the diet of post-metamorphic frogs postulates that it happens more often in bigger frogs. Moreover, this hypothesis has yet to be tested. Methods. We reviewed the literature on frog diet in order to test the size hypothesis and determine whether there are other putative explanations for anurans in the diet of post-metamorphic frogs. In addition to size, we recorded the habitat, the number of other sympatric anuran species, and whether or not the population was invasive. We controlled for taxonomic bias by including the superfamily in our analysis. Results. Around one fifth of the 355 records included anurans as dietary items of populations studied, suggesting that frogs eating anurans is not unusual. Our data showed a clear taxonomic bias with ranids and pipids having a higher proportion of anuran prey than other superfamilies. Accounting for this taxonomic bias, we found that size in addition to being invasive, local anuran diversity, and habitat produced a model that best fitted our data. Large invasive frogs that live in forests with high anuran diversity are most likely to have a higher proportion of anurans in their diet. Conclusions. We confirm the validity of the size hypothesis for anurophagy, but show that there are additional significant variables. The circumstances under which frogs eat frogs are likely to be complex, but our data may help to alert conservationists to the possible dangers of invading frogs entering areas with threatened anuran species.