There have been very few studies that have describe the epidemiology of first-time venous thromboembolism (VTE) in a large, ethnically diverse population. The California Discharge Data Set was used to identify a cohort of cases with incidentVTE in 1996. Cases associated with traditional provoking risk factors were identified and the remaining cases were labeled as idiopathicVTE. Direct standardization using census information was performed to compare incidence rates across races, gender, and gender within race. There were 21,002 cases with incident VTE in 1996, a crude incidence of 90 events per 100,000 adults. Thirty percent of all VTE events were pulmonary embolism. The directly standardized incidence per 100,000 California adults was 93+/-1.7 (+/-95% CI) in women, 85+/-1.7 in men, 103+/-2.1 in Caucasians, 138+/-6.5 in African- Americans, 61+/-2.8 in Hispanics and 29 +/- 2.4 in Asian-Pacific Islanders (p<0.001 for all inter-group comparisons). After adjusting for misclassification of race, the incidence of VTE per 100,000 was 104 in Caucasians, 141 in African-Americans, 55 in Hispanics, and 21 in Asian/Pacific-Islanders. The incidence of idiopathic VTE was significantly lower among both Hispanics and Asian/Pacific-Islanders (p<0.001) than Caucasians or African-Americans. African-Americans were more likely, and Hispanics less likely, to be diagnosed with idiopathic pulmonary embolism compared to Caucasians. The 28 day case-fatality rate among cases with idiopathic VTE was 2%, and it was significantly higher among African-Americans (4.1%) compared to Caucasians (1.8%, p<0.001). There are important differences in the incidence of total and idiopathicVTE and in the proportion of events diagnosed as pulmonary embolism among each of the major racial/ethnic groups in California. Further research is needed to explain these observed differences among the different racial/ethnic groups.