Marine systems experience an unprecedented number of stresses caused by humans. Over the last 25 yr an increasing amount of attention has been given to examining the combined impacts of multiple stressors. Yet, existing studies reveal few patterns that facilitate predicting or understanding when multiple stressors should combine additively, synergistically, or antagonistically. One contributing factor to this lack of clarity may be the lack of a common framework that is based on a mechanistic understanding of stressor impacts. We adapt and advocate a general framework that is employed by the US EPA in terrestrial systems for use in marine systems. This framework involves 3 steps: (1) Mechanistically examine the impacts of multiple stressors on individual organisms. (2) Scale these impacts on individual organisms to population level responses to multiple stressors. (3) Examine context-dependent changes in stressor responses due to changes in community or ecosystem properties. We also argue that 3 specific aspects of previous studies hamper our ability to detect patterns in multiple stressor impact. First, a large number of studies have reported impacts on growth, survival, etc., without elucidating mechanisms. Second, the majority of studies provide insufficient data to determine whether threshold or nonlinear responses to stressors occur. Third, 32% of existing studies transformed data to meet model assumptions, but in so doing, they unknowingly altered the statistical model being tested. We argue that rectifying these 3 conditions will accelerate the detection of patterns in the way that multiple stressors combine to influence marine systems.