Background: Searching for and selecting studies according to Cochrane quality criteria can be cumbersome. A range of bibliographic databases should be used, with highly sensitive search strategies but relatively low precision. This results in large numbers of references which then must be examined by several reviewers independently. These tasks currently require the use of several softwares and keeping track of the whole process is uneasy. Objectives: We aimed at developing a unique interface to facilitate the management of references and the study selection process by several reviewers. Methods: We developed a web service which allows importing the references identified from the major databases (including CENTRAL, Medline, Embase, Lilacs, Science Citation Index). Users can input the search strategies and dates of search for each database. We implemented an algorithm to automatically identify duplicate records of the same report. The system also allows linking together multiple reports of the same study. Multiple user accounts allow reviewers examining references independently. Step-by-step assessment of titles and abstracts with a consensus procedure for disagreements is handled. Measurements of agreement are produced at each step. Finally, the numbers of hits retrieved by the electronic searches and the numbers of selected records are automatically documented in a flow chart of the whole study selection process. The web service was developed using the Python Programming Language. Results: So far, the web service is operational. An evaluation study is underway, involving physicians and librarians. We expect that the use of this tool will result in saving time and increasing the accuracy of study selection. Results will be presented at the conference. Conclusions: We expect this free web service a unique interface to perform the whole selection process- to be a helpful tool for Cochrane reviewers.