Variations in health service use rates by geographic area have long interested researchers and policymakers. Typically, investigators comparing population-based health care utilization rates among geographic areas have demonstrated substantial variations in use among seemingly similar communities. One method of investigation is "small area analysis." Numerous areas in North America have been studied extensively using this technique. This research has attempted to document the amount of variation found in health care use rates among areas; determine whether or not there is a pattern to such use in high- versus low-use areas; and identify the variables that are associated with the variation and explain a portion of the variation. Beyond this, many researchers have attempted to ascertain whether such variables are associated with characteristics of the population, whether they reflect differences in access and need, or whether a substantial portion of the variation is associated with differences in the medical care system itself. This review discusses the methods used to define the areas, the dependent variables that have been studied and the patterns found within them, the independent variables that have been tested, the statistical methods and analysis procedures used, the results of each study, and the policy recommendations emanating from the research. More importantly, based on what has been learned, the paper provides researchers in small area analysis with a set of recommendations for both analyzing and reporting results. These recommendations are designed to facilitate the development of a common research methodology, increase the comparability across studies, and enhance the use of this technique in the health policy formulation process.