While normalized bibliometric indicators are expected to resolve the subject-field differences between organizations in research evaluations, size still matters. Furthermore, research organizations, policymakers and research funding providers tend to use benchmark units as points of comparison for a given research center in order to understand and monitor its development and performance. In addition to monitoring and evaluations, the identification of comparable benchmark organizations can also be used to pinpoint potential collaboration partners or competitors. Therefore, methods to identify benchmark research units are of practical significance. However, few studies have investigated this problem. This study aims to propose a bibliometric method to identify benchmarks. We define a benchmark as a well-connected research environment, in which researchers work on similar topics and publish a similar number of publications compared to a given research center during the same period. Three essential attributes for the evaluation of benchmarks are research topics, output, and coherence. We apply this strategy to a Swedish research center, and examine the effectiveness of the method.