Use of cover objects for hiding is a well developed behavior in reptiles, including snakes. Snakes sometimes aggregate under cover objects, which may reflect a shortage of suitable cover, very favorable conditions, social attraction, or simply chance. However, most studies of aggregation behavior have been conducted in the laboratory. In this study, I investigated the tendency of six species of snakes to aggregate under rocks in the field in southern Ontario, Canada. Most snakes under rocks were by themselves, but I found single- and mixed-species aggregations of up to four snakes. Although all species were involved in mixed-species aggregations, I did not find all possible combinations of species; no aggregations involved more than two species. Goodness-of-fit tests suggested that the pattern of aggregation sizes was well described by a geometric distribution, implying a nonrandom tendency toward aggregation. Nonetheless, because data were pooled over time, random occurrence of aggregation cannot be ruled out. However, aggregations were clearly nonrandom in another sense: individuals in aggregations tended to be the same size, perhaps indicating size-specific choice of rock and/or fellow snake. Careful field experimentation will be required to test these ideas.