Two separate field experiments were performed in the U.S.A. and Norway with stocked Atlantic salmon Salmo salar. In the Norwegian experiment, the offspring of early-spawning fish which had larger eggs and emerged a few days before offspring of later spawning fish had consistently higher survival rates. In the U.S.A. experiment, stream sections with higher proportions of favourable foraging locations during the critical period (the transition from dependence on maternally-derived yolk reserves to independent feeding) had lower loss rates of fish stocked as unfed fry. These results provide support for the critical period concept (CPC) in Atlantic salmon, underscores the utility of a manipulative approach to achieve further advances in knowledge of Atlantic salmon ecology and provide additional guidance to management and restoration. A mechanistic, conceptual model for density dependence is presented to identify important knowledge gaps that remain to further evaluate the importance of the CPC for Atlantic salmon population regulation.