Kin selection theory predicts conflict in social Hymenoptera between the queen and workers over male parentage because each party is more closely related to its own male offspring. Some aspects of the reproductive biology of the bumble–bee Bombus terrestris support kin selection theory but others arguably do not. We present a novel hypothesis for how conflict over male parentage should unfold in B. terrestris colonies. We propose that workers delay laying eggs until they possess information showing that egg laying suits their kin–selected interests. In colonies where queens start to lay haploid eggs early, we hypothesize that this occurs when workers detect the presence of queen–produced male brood in the brood' s larval stage. In colonies where queens start to lay haploid eggs late, we hypothesize that it occurs when workers detect a signal from the queen to female larvae to commence development as queens. Our hypothesis accounts for previously unexplained aspects of the timing of reproductive events in B. terrestris, provides ultimate explanations for the results of a recent study of mechanisms underlying queen–worker conflict and helps explain this species' characteristic bimodal (split) sex ratios. Therefore, kin selection theory potentially provides a good explanation for reproductive patterns in B. terrestris.