Background: Within the framework of a systematic review, a critical task is the reliable selection of potentially relevant literature from a comprehensive pool of references stored in a database. Objectives: To facilitate the literature selection process, we developed and tested a new software tool. Methods: We developed and tested this tool within the framework of several systematic reviews conducted by the German Institute for Quality and Efficiency in Health Care. The technology is based on a Visual Basic for Applications (VBA) EXCEL macro. The interface for Reference Manager or Endnote is provided by tag-format import / export facilities. The tool covers the entire process, starting with the title / abstract screening, to obtaining and evaluating potentially relevant full-text publications, to the documentation of the selection process as a whole by one or more independent raters. One key feature of the tool is the support of the screening process by highlighting individually defined keywords within specific data items. This should help to ensure the selection of relevant and the preclusion of irrelevant references. The applied set of key words, which can be organized by colour and font type, will be refined in an ongoing learning curve. The tool further supports the merging of separate databases (e.g. processed in parallel by several raters) and the compensation of inter-rater differences. Results: A literature search for systematic reviews on positron emission tomography (PET) found 6533 datasets, corresponding to 3660 pages (2500 letters per page) that had to be screened. The tool s reduction in effort, compared to manual screening, was estimated as being 44% (translating into 1600 pages less), depending on the length of the datasets (15% reduction for every multiple of 500 letters). This model is likely to estimate the real reduction in effort, as experienced in the PET project. Conclusions: Using this software tool in the literature selection process may save resources and improve precision.