Two main hypotheses proposed to explain the seasonal decline in reproductive performance in birds are (1) deterioration of environmental conditions and (2) lower parental quality of late breeders. Previous experimental work addressing these hypotheses generally have problematic biases pertaining to delay of hatching, costs of re-laying and incubation, as well as variation in the quality of eggs, territories, offspring and parental traits. We address these biases in an experimental test of the timing hypothesis (i.e. (1) above) in a precocial bird. Using a 2-year cross-over design and game-farm mallard (Anas platyrhynchos) eggs originating from a number of hens and a standardised delay procedure, we introduced early and late broods with a foster female onto boreal oligotrophic lakes and monitored subsequent duckling survival. Standardised invertebrate sampling was done concurrently to get a measure of lake-level abundance of aquatic prey, a likely causative agent of putative seasonal difference in duckling survival. Survival data and covariates (duckling age; days) were analysed by an information theoretic approach. There was no effect of treatment (i.e. manipulation of hatching date) on duckling survival, which was higher in 2005 than in 2004. In contrast to observational studies from more seasonal wetlands, our experiment demonstrates that duckling survival on boreal lakes was not affected by a 12-day delay in hatching date. Since we did not find any consistent trends in abundance of aquatic prey, i.e. neither clear peaks nor differences between treatment periods, we hypothesise that moderate climate change has minor effects on resource abundance and hence also on mallard duckling survival in boreal environments.