We conducted the current study to compare the incidence and risk factors of arterial thrombosis in lupus and non-lupus primary glomerulonephritis. We identified patients in whom lupus nephritis and non-lupus primary glomerulonephritis were diagnosed between 1993 and 2003 using our lupus cohort database and pathology registry. We analyzed the cumulative incidence of new arterial thromboembolic events since diagnosis by Kaplan-Meier plot, and studied risk factors by multivariate analysis. We studied 162 patients with lupus and 181 patients with non-lupus primary glomerulonephritis. After a mean observation of 8.1 years, 47 (14%) patients died, 23 (7%) were lost to follow-up, and 38 (11%) developed 42 arterial events (incidence, 15.1/1000 patient-years). Although patients with lupus nephritis were younger and had a significantly lower frequency of smoking, hypertension, obesity, and renal dysfunction, their cumulative risk of arterial event at 5 years was not significantly lower than that of patients with primary non-lupus glomerulonephritis (6.3% vs. 6.6%, p = 0.96). In a Cox regression model, lupus was found to be an independent risk factor for arterial thrombosis (hazard ratio 3.57 [1.07-11.9]; p = 0.04), in addition to increasing age (hazard ratio 1.04 per year; p = 0.02), low-density lipoprotein > or =2.6 mmol/L (hazard ratio 4.46; p = 0.002), and glomerular filtration rate <30 mL/min (hazard ratio 2.67; p = 0.04). We concluded that in immune-mediated glomerulonephritis, having systemic lupus increased the risk of arterial thromboembolism after adjustment for age, renal insufficiency, and other traditional risk factors.