Although studied for 35 years, knowledge of the reproductive biology of the King Penguin Aptenodytes patagonicus remains incomplete. The chick requires more than 12 months of care, which extends the breeding cycle, including moult, to more than one year, i.e. the King Penguin is neither annual nor biennial. In an attempt to resolve ambiguities in the literature and to elucidate the long-term breeding strategy of the species, we studied breeciing frequency at the individual level, considering the decision to breed in relation to breeding history over the previous few years. Although adult birds attempted to breed annually (0.83 breeding attempts per year), successful rearing occurred, at best, every two years only (maximum of 0.41 fledged chick per pair). Comparing successive years, the number of breeding birds in the colony was stable but the number of fledged chicks varied from 29 to 278 over eight years. These results suggest that King Penguins adopt (as individuals) an opportunistic reproductive strategy, in that they usually lay an egg every year, even when failure is certain. Nevertheless, the decision to breed was not entirely blind, and we identified groups of birds that invested differentially in breeding attempts. The decision to breed was related to the previous breeding frequency, i.e. 81% of the birds that had bred continuously in the past started a new breeding attempt, but only 67% of birds that had missed a year did so. In intermittent breeders, birds that had bred frequently, more often started a new breeding attempt than birds that had largely missed years (71% versus 57%, respectively). Classes of breeders could correspond to age classes, to birds of different breeding quality or to alternative breeding strategies coexisting in the species. Testing the hypothesis that reproductive effort increases with age should be possible in future.