When players compete repeatedly, prizes won in earlier contests may improve the players’ abilities in later contests. This paper determines the allocation of prizes within and across contests that maximizes the (weighted) sum of aggregate efforts.