Monitoring is important to assess the effectiveness of management, however for monitoring to be effective it is critical to understand if the monitoring design used will likely detect population changes, should they occur. Will the data collected have enough statistical power? For four forest bird species of conservation concern, we evaluated the statistical power of alternative monitoring designs on Christmas Island to detect population declines over a 10-year period. We combined estimates of detection probability with species distribution models to predict the effectiveness of alternative scenarios for monitoring design. For each scenario, we modified the survey effort in terms of the number of sites visited, number of visits per year, and number of years between surveys. We identified both new priority areas for surveys and which survey variables contributed the most to detecting population declines.