Natural selection may influence the arrangement of individuals into mated pairs through either inter-sexual (mate choice) or intra-sexual selection (competition). A study of the American kestrel, Falco sparverius, in northern Saskatchewan distinguished between these two processes using size as a measure of the bird's competitive ability, and condition (mass scaled to body size) as an index of quality. Both sexes arrive on the study area after spring migration in equal numbers and males establish territories. Males and females that moved among territories at the time of pair formation were not different in size or condition from those that did not move, suggesting that birds were not being displaced by superior competitors, and that females moved to encounter potential mates. Within mated pairs, there was no relationship between a bird's size and the condition of its mate for either sex as would be predicted if intra-sexual competitition explained mating patterns. Instead, there was positive assortative mating by condition, suggesting that both sexes used quality as the criterion in choosing mates. There was no correlation between the sizes of males and females in mated paird. Because there were no differences in size or condition of breeding and non-breeding males, factors other than physical attributes, such as prior experience with the area, may determine a male's success in obtaining a territory. Because females that did not obtain mates were in poorer condition than those that did, males may have rejected poor quality females. The results suggest that intra-sexual competition was not important for pair formation, and that kestrels chose mates on the basis of quality.