To test for the existence of a reproductive cost, we manipulated brood sizes (-2 and +2 nestlings) over 6 years in a northern population of Willow Tits Parus montanus breeding in natural holes. Possible effects were sought in subsequent survival and fecundity of the parents. Parents given extra chicks made more feeding visits than did parents with reduced and control broods. However, this was not reflected in differences in parental body-weight between groups at the end of the nestling period. Brood size manipulation did not significantly affect female or male survival. In 4 out of 6 years, there was a weak and nonsignificant effect on male survival, consistent with a cost of reproduction. Female and male fecundity in the year following the experiment was not affected by the manipulations. Thus, the data do not give evidence of an intragenerational cost of reproduction in the Willow Tit. Parents appeared unwilling to increase their breeding effort to a level which jeopardized their own survival or future breeding success. It is possible that, because of the time constraints in northern latitudes, females work under their capacity and lay smaller clutches than would otherwise be most profitable. Thus, no costs to the parents would be expected as a consequence of manipulations. These results suggest that the current reproductive rate is not maintained by reproductive cost in the Willow Tit. However, the results do not rule out the possibility that selection has operated outside the current range of reproductive rates during evolutionary history of the species.