In crested penguins (Eudyptes spp.), second-laid eggs typically hatch before first eggs. Amongst a variety of factors that have been considered as mechanisms underlying this reversal, has been the idea that crested penguins can adjust the degree of hatching asynchrony by manipulating egg positions (i.e. placing the smaller first egg in the supposedly thermally disadvantaged anterior position) during incubation (termed Preferential Incubation Hypothesis). We tested this in the Snares crested penguin (Eudyptes robustus) and the closely related, but synchronously-hatching, yellow-eyed penguin (Megadyptes antipodes). Snares crested penguins were more likely to place their first eggs, which are smaller than second eggs, in the anterior incubation position than were yellow-eyed penguins, which have a clutch of two similar-sized eggs. But when yellow-eyed penguins, a non-brood reducing species, were provided with an artificial size-dimorphic clutch, they also placed smaller eggs more frequently in the anterior position, suggesting that a general preference exists among penguins to place smaller eggs in the anterior position. Egg temperatures of small first eggs of Snares crested penguins were higher in the anterior than in the posterior position. Large first eggs in lesser size-dimorphic clutches experienced high temperature differences in relation to position, while small first eggs in greater size-dimorphic clutches were incubated at similar temperatures. In yellow-eyed penguins, large eggs within clutches generally had higher egg temperatures than small eggs. Incubation periods of second eggs declined with increasing egg size. Egg-size variation, rather than egg positioning behaviour, influenced hatching patterns in Snares crested penguins. In lesser size-dimorphic clutches, second eggs were more likely to hatch first while in greater size-dimorphic clutches, small first eggs were more likely to hatch at the same time or before the second eggs. This was similar in yellow-eyed penguins, where second eggs hatched earlier in clutches with large first eggs. Our data contradicts the Preferential Incubation Hypothesis and we conclude that this hypothesis is unlikely to explain the reversed hatching asynchrony in crested penguins.