Nine oat cultivars and experimental lines from four germplasm sources were crossed in a diallel mating design without reciprocals. F(,1) heterosis for grain yield was evaluated in two experiments, and 48 F(,2)-derived lines from each of the 36 matings were evaluated for bundle weight, grain yield, straw yield, harvest index, height, and heading date in two experiments. Number of transgressive segregates per trait and generalized genetic variance were calculated for each mating. Four measures of genetic distance between the parents were calculated: geneological distance, Euclidean distance based upon principal components, and the distance measures proposed by Hanson and Casas and Cervantes et al. The relationships between the four distance measures and the three types of breeding behavior were examined via correlation and regression. Correlations, where significant, were low to moderate. Regressions, where significant, were primarily linear with low R('2) values. The regression of heterosis on Euclidean distance in one experiment was quadratic with a high R('2) value. Using information from more than one distance measure improved the R('2) values for polynomial regression.