Theoretical and empirical studies predict that there should be a decrease in hunting success of predators with increasing prey group size. Most of these studies investigated situations in which predator and prey were in full view of each other before, during and after an attack. In this study, single rock bass Ambloplites rupestris were given an opportunity to launch surprise attacks at shoals of creek chub Semotilus atromaculatus that ranged in size from two to 13 fish. There was no significant influence of either shoal size or attack distance on predator success rate and no significant relationship between attack distance and shoal size. Furthermore, it was found that the leading fish of a shoal was attacked significantly more often than fish in other shoal positions, indicating that predation risk was not shared equally among shoal members. Also, leading fish in larger shoals (eight to 13 fish) were not more likely to survive a predator attack than ones in small shoals (two to seven fish).The consequences of these results are discussed in the general context of antipredator benefits of grouping.