The sustainability of exploitation is based on density-dependent renewal of populations: when population density decreases as some individuals are taken, the remaining individuals compensate by surviving or reproducing better. In general there is a trade-off between two desired outcomes: a high yield and a high remaining population size. A hunting strategy is Pareto optimal if it balances this trade-off without wasting possibilities of improving the performance in either aspect. Lack of knowledge concerning the age structure, mating system or density dependence operating in a population will very easily cause suboptimality in this sense, whereas utilising knowledge of density dependence may, in some cases, even overcome the conflict between the goals, so that harvesting can increase rather than decrease population sizes. Suboptimal timing of harvesting is an example which not only causes unnecessary harm to a population, but also hampers estimation of the compensatory or additive nature of mortality. A bias towards additivity will be found if hunting and natural mortality overlap in time, and even ‘superadditive’ results are possible. A mortality pattern that appears additive cannot, therefore, be used to deduce that overwinter survival is density independent. These results have consequences to harvest planning. Adjusting the length of the open season is a tool frequently used to regulate the harvest. Since estimated slopes of compensation cannot be assumed to remain constant if the timing of the open season is changed, the effect of a prolonged season will be more drastic than a mere change in kill rates would predict. Such factors are likely to have the strongest effects in species with long harvest seasons, such as many migratory European waterfowl.