One goal in indicator development is to implement long-term monitoring that will track the relative condition of the indicator over time. Among the first steps in establishing a monitoring program is to develop a sampling design that adequately characterizes the indicator to be monitored as well as the cost-effectiveness of the program. We used breeding bird data collected in Lake Superior and Lake Michigan coastal wetlands (riverine, lacustrine, barrier-protected) to determine: 1) how to select individual wetlands for sampling, 2) optimum number of sample points per wetland, 3) optimal daily sampling period, 4) how many times to sample, and 5) the costs associated with implementing a monitoring program for breeding bird communities of wetlands across the Great Lakes. We found that wetlands selected for sampling should represent the range of wetlands sizes available for monitoring and that the most cost-effective strategy would be to sample a maximum of three points, even in the largest wetlands. Because surveys conducted in the morning recorded a much higher (P < 0.001) number of species and individuals, we recommend that morning surveys should be conducted. Increasing number of wetlands sampled should be the first priority because sample precision is more improved at a higher cost ratio than by adding counts to the same wetland. Multiple visits to wetlands should be considered only after maximizing the number of individual wetlands visited with money available for surveys. We calculated that the average costs would be approximately 50.00 USD/year (2001 dollars) to monitor one wetland using one morning survey for breeding birds.