Aquatic ecosystems are confronted with multiple stress factors. Current approaches to assess the risk of anthropogenic stressors to aquatic ecosystems are developed for single stressors and determine stressor effects primarily as a function of stressor properties. The cumulative impact of several stressors, however, may differ markedly from the impact of the single stressors, and can result in non-linear effects and ecological surprises. To meet the challenge of diagnosing and/or predicting multiple stressor impacts, assessment strategies should focus on properties of the biological receptors rather than on stressor properties. This change of paradigm is required because (i) multiple stressors affect multiple biological targets at multiple organizational levels, (ii) biological receptors differ in their sensitivities, vulnerabilities and response dynamics to the individual stressors, and (iii) biological receptors function as networks, so that actions of stressors at disparate sites within the network can lead via indirect or cascading effects, to unexpected outcomes.