The stability of causal attribution over repeated test days was examined using a 2 (win, loss) X 2 (immediate, one day delay) X 3 (post game, 2 weeks, 4 weeks) design. Male high school football players outlined the degree of causality attributed to ability, effort, luck, task difficulty and officiating. The results from an ANOVA analysis supported a view that attributions are stable over a four week interval. However, a correlational analysis revealed that there is only a moderate to low consistency of individual differences in attribution over repeated tests. Thus, retrospective judgements for causality may not be reliable.