Recent attention has focused on the role that sperm competition may play in the evolution of sperm morphology. Theoretical analyses predict increased sperm size, decreased sperm size, and no change in sperm size in response to sperm competition, depending on the assumptions made concerning the life history and function of sperm. However, although there is good evidence that sperm morphology varies widely within and between species, the adaptive significance of this variation has not been examined. Here we document significant intraspecific variation in sperm length in the field cricket, Teleogryllus oceanicus. Sperm length did not influence the rate of migration of sperm from the spermatophore to the female's spermatheca. We performed sperm competition trials in which we varied the numbers of sperm transferred by each of two males that differed in the length of sperm they produced. Neither sperm length nor the number of sperm transferred influenced paternity. The same results were obtained using two different methods for assigning paternity. The distribution of paternity across a female's mates was highly variable, with frequently one, or more in the case of females mated to four males, principal sire. There were no mating order effects on paternity. These data show that sperm do not mix randomly in the female's spermatheca. We discuss several alternative explanations for the patterns of paternity observed.