Attempts to obtain field data for risk assessment of contaminants released into marine/estuarine systems can be complicated by a number of interrelated factors, such as complex circulation and mixing patterns, diverse stratification forces, dynamic short-term changes as well as seasonal movements of biota, and the ecosystem’s physical scale. Tests conducted in simulated ecosystems are subject to constraints that restrict the effect of physical forces, limit physical scale of the test, and introduce biases from chemical partitioning and processing along the walls of the test system. These constraints restrict the broad application of test results as a model of dynamic marine systems. Through selected examples from literature and ongoing studies, we provide illustrations of how contaminant effects are studied at the individual, population, and community level in the field and/or in simulated ecosystems, such as mesocosms. We discuss marine-environment field studies and simulated field studies that measure contaminant effects with respect to exposure-response relationships, food-web interactions, competition/colonization studies, and selected aspects of nutrient cycling. Based on results to date, we conclude that: (1) successful field studies must focus on selected endpoints fundamental to our understanding of contaminant effects, and (2) endpoints studied in simulated ecosystems must be representative of key structural and/or functional factors of the system of interest.