Seasonal variation in reproductive success is a common feature of most organisms. To understand the evolution of breeding seasons and reproductive strategies of individual animals, it is necessary to assess the extent to which seasonal variation in reproductive success is causally related to seasonal variation in the environment ('timing' hypothesis), to differences in quality between early and late breeders or their territories ('quality' hypothesis), or to a combination of both. We manipulated timing of breeding in the Great Tit Parus major, a small passerine, to test these hypotheses. A group of experimentally delayed birds was created by removing first clutches, inducing birds to lay a replacement clutch. Reproductive success of delayed pairs was compared with control pairs that bred early and with pairs that bred late. We conclude that seasonal declines in reproductive success at the nestling stage and survival of adult females were caused by differences in quality between early and late breeders. Recruitment of fledglings into the breeding population and the occurrence of second clutches were causally related to the timing of breeding. The seasonal decline in clutch size was caused by a combination of timing and quality effects. We attempted to assess the relative importance of timing and quality in the seasonal decline in reproductive success, expressed as lifetime production of recruits. We tentatively conclude that 87% of the seasonal decline in lifetime reproductive success could be attributed to a timing effect per se, whereas quality differences between early and late breeders accounted for the remaining 13%.