Various kinds of no-take areas (refuges, reserves) are gaining attention as conservation tools. The efficacy of reserves can be considered from the perspectives of providing baseline data sets, protecting the stock, maximizing yield to the fishery, or some combination of these. Regardless of the measure of effectiveness of a reserve, practical application requires the development of techniques for settling operational and policy questions such as how large a reserve should be. A simple model, involving population growth and harvest, is used to explore how the fraction of habitat assigned to a reserve affects the sustainability of a take and to frame the trade-off between control of harvest outside of the reserve and the size of the reserve. This exploration also leads to the discovery of a robust conservation invariant for reserves.