Present-day ecosystem management involves understanding of the synergistic effect of multiple stressors on multiple and frequently nebulous management end-points. An example is the simultaneous management of nutrient load reductions and salmon stocking in Lake Ontario. In this study, a simple whole-lake annual time scale model was developed to assess the relationship between these two stressors and various ecosystem responses. The model was used to explore the utility of some possible management end-points for ecosystem health. In historical simulations, production per stocked fish and salmon survival appeared to be good indicators, while nutrient recycling rate and average ecosystem-wide food limitation were found to be fairly unresponsive to the two stressors. The model was further used to predict long term averages of salmon biomass and selected health indicators at various sustained loading and stocking rates. Salmon biomass increased with stocking rate at all stocking rates examined, but the rate of increase declined somewhat at high stocking rates. The response of salmon biomass to nutrient loading appeared to be approximately sigmoidal i.e. there was a nutrient threshold below which fish biomass could not be sustained and another nutrient threshold above which salmon biomass either remained constant or even decreased. The response to either stressor was found to be modified by the value of the other stressor, illustrating the importance of ecosystem-level models for aquatic ecosystem management.