Resource managers must often act to protect fisheries and fish habitat without the certainty that their actions are justified. Delivering the science needed to support and direct management decisions is a daunting exercise, likely beyond the capabilities of a single research group or management agency. This problem is exacerbated by the lack of a common framework to formulate and test important hypotheses about biotic response to aquatic habitat change. A partial solution may be provided by co-operative research networks to produce an integrated design and synthesis of quasi-independent studies within a common framework for hypothesis generation and testing. A well-designed framework should attract scientists and agencies who recognize the benefit of cooperative research. We demonstrate such an approach by using it to test hypotheses about lake fish community response to habitat change. Our framework includes a list of hypotheses, a list of treatments (i.e., habitat manipulations), an experimental design specifying the number of lakes per treatment, and advice for measuring habitat and fish parameters. Because our procedure uses ‘before-after’ comparisons to measure effects of habitat changes, lakes can be studied independently (and hypotheses can be tested independently) yet still contribute synergistically to the larger experiment. A ‘staircase’ design, ensuring that treatment effects are independent of environmental correlates such as climate variables, would be implemented, largely by default, because contributions to the design would accumulate over time. We believe this cooperative approach will improve the ability of researchers to meet the growing demands for useful, reliable aquatic science.