<p><b>PURPOSE: </b>The discipline of family medicine (FM) lacks a comprehensive methodology, which can be applied as a standard for assessing overall research output in both the field of FM and by general practitioners (GPs)/general practice institutions. It was the aim of this study to develop a sensitive search strategy for assessing publication output in the field of FM independent of the author's profession or affiliation and by GPs/general practice institutions independent of their field of scientific interest.</p><p><b>METHODS: </b>Literature searches limited to the year 2005 were conducted in PubMed and ISI Web of Sciences (ISI WoS). In PubMed, all relevant MeSH terms were used. Search terms possibly contained in the author's affiliations have been collected. In ISI WoS, the same entry terms including their abbreviations and plural forms were applied. The final queries were validated by manual review and matching results with selected FM journals.</p><p><b>RESULTS: </b>A comprehensive list of combined search terms could be defined. For the field of general practice/FM more publications could be retrieved in PubMed. Almost twice as many publications by GPs/general practice institutions could be retrieved in ISI WoS, where--in contrast to PubMed--the affiliation is documented for all authors.</p><p><b>CONCLUSIONS: </b>To quantitatively assess publication output in the field of FM, PubMed was identified as the preferable database. To assess publication output by GPs/general practice institutions, the ISI WoS is recommended as the preferable database. Apparently, the ISI WoS is more suitable to compare the research productivity of different countries, authors or institutions.</p>