Aim: The aim of this systematic review was to analyse the effects of gatekeeping where primary care physicians (PCP) control access to specialist care. Methods: Literature search in Medline, EMBASE, Cochrane Library, and a hand search were carried out. Inclusion criteria: (1) intervention: gatekeeping by PCP compared to free access to specialist care; (2) outcomes: health outcomes, health related quality of life, quality of care, utilization of care, costs, satisfaction of patients and providers; (3) design: RCT, quasi-random. CT, CBA, cohort and case control studies, ITS. Data extraction and assessment was done by two independent reviewers according to Cochrane EPOC-Group and USTFCPS. Results: 24 included studies (1989-2007) were as follows: 1 RCT, 2 quasi-randomised CT, 3 prospective, and 12 retrospective cohort studies, 4 CBA, and 2 ITS. 67 % of the studies analysed data from the USA, the remaining from CH, UK, DK and NL. Studies had relevant limitations concerning the quality of execution and publication. Overall 13 of 24 studies reported a positive and two a negative effect of gatekeeping compared to open access models; nine showed no differences. The results varied according to outcome parameters. Conclusions: International evidence on effects of gatekeeping is limited by the low internal validity of studies and applicability to other contexts. It suggests that gatekeeping by PCP decreases utilization of specialist care and health care costs. Based on very few studies health outcomes and patient quality of life in gatekeeping models might be comparable with those in open access models. Evidence is inconsistent or not available concerning the quality of care, patient or provider satisfaction.