Sex investment ratios in populations of bumblebees are male biased, which contradicts theoretical predictions. Male–biased investment ratios in eusocial Hymenoptera are assumed to be non-stable for both the queen and her workers. In this paper, we show that male–biased sex allocation does not necessarily decrease fitness in the bumblebee Bombus terrestris. A male–biased investment ratio can be the result of an optimal allocation of resources when resources are scarce if (i) there is a large cost difference between male and female production, (ii) there is uncertainty about the amount of resources a colony can invest, and (iii) only a proportion of the investment made in an individual can be reused. This resource allocation then leads to split sex ratios depending on the amount of resources available to a bumblebee colony: colonies under low resource conditions will show a male–biased investment ratio, whereas colonies under high resource conditions allocate more resources towards females. However, the extent to which bumblebee populations show a male–biased sex allocation cannot be explained by cost differences between male and female production alone. In a recent paper, A. F. G. Bourke argued that male–biased investment ratios in bumblebee populations are a by–product of the occurrence of protandry (males emerge before females). Here we will extend Bourke's argument and show that within a protandrous population, both protandrous and protogynous (females emerge before males) colonies exist. The existence of protandrous and protogynous colonies results in split sex ratios in time, because protogynous colonies rely on males produced by protandrous colonies (partial protandry).