To evaluate the possibility that fibrinogen represents a cardiovascular risk factor. A computerized literature search (1980 to 1992) identified all published epidemiologic studies on fibrinogen and cardiovascular disease. Clinical and basic research data were found by separate searches. References of all papers thus obtained were studied and relevant papers included. Six prospective epidemiologic studies were included in a meta-analysis (one study was excluded because the study population was nonrepresentative). Clinical papers were reviewed separately for other evidence of causation. The correlation of fibrinogen levels on the subsequent incidence of myocardial infarction, stroke, and peripheral arterial occlusive disease was assessed and the causality of the association was analyzed. Calculations were made to examine fibrinogen level (in tertiles) versus cardiovascular risk. Odds ratios of high versus low tertile were computed. All prospective studies showed that fibrinogen was associated with subsequent myocardial infarction or stroke. A total of 92 147 person-years was covered by these investigations. Odds ratios varied between 1.8 (95% CI, 1.2 to 2.5) in the Framingham and 4.1 (CI, 2.3 to 6.9) in the GRIPS study, with a summary odds ratio of 2.3 (CI, 1.9 to 2.8). Associations existed between fibrinogen and other cardiovascular risk factors, but after multivariate analysis, only the association between fibrinogen and cardiovascular events remained. The majority of the preconditions for causality were fulfilled, indicating that fibrinogen is pathophysiologically related to cardiovascular events. Fibrinogen can be considered a major cardiovascular risk factor. Future studies of cardiovascular morbidity and death should include this variable.