The ecological assessment of long term or reproductive effects from pesticide exposure in birds and mammals is currently seen to be problematic. In birds especially, the current test results are difficult to extrapolate to a field situation. For example, whereas a majority of laboratory studies report clutch size reductions in response to graded pesticide doses, this effect is rarely, if ever, seen in the wild in response to exposure to toxicants. Also, current laboratory tests in birds provide a very truncated measure of reproductive performance. At the opposite end of the spectrum, the ecological assessment of long-term impacts to wild mammal populations suffers from an overabundance of very detailed information generated on a few rodent and non-rodent test species. It becomes critical here to identify which of the multitude of assessment endpoints should form the basis of an ecological assessment. Whether bird or mammal, I suggest that it is essential to separate parental from developmental endpoints and to apply the different effect levels to agent-based models of population growth and performance. Only then can we hope to gain a real understanding of the true long-term consequences of pesticide or other chemical exposure.