In the present literature there is still controversy as to whether patients with systemic lupus erythematosus (SLE) are at increased risk of developing malignant diseases. In recent years a number of epidemiological studies have been conducted and some have suggested an association between SLE and malignant diseases while other studies have not. The objective of this study was to investigate this relationship in an unselected cohort of Icelandic patients with SLE. All patients diagnosed with SLE registered in the Icelandic SLE database were compared to the Icelandic cancer registry. For completeness, hospital charts and outpatient notes were also reviewed. The study period was from 1957 to the end of 2001. The O/E (observed/expected ratio), CI and P-value were calculated for total number of malignancies as well as individual malignancy types. Of 238 patients diagnosed with SLE (213 women and 25 men) 39 malignancies were diagnosed in 36 patients; 32 women and four men. Of the 36 patients, 27 were diagnosed subsequently with SLE and malignant disease. The mean age at diagnosis of SLE was 43.2 years (range 10-81) and at time of diagnosis of malignancy 62.7 years (range 43-86). The O/R for the whole study population was 1.38 (CI 0.89-1.87, P = 0.063), 1.45 for the women (CI 0.91-1.99, P = 0.051) and 1.03 for the men (CI 0.22-2.66, P = 0.560). The O/R for the most frequent malignancies diagnosed subsequently to SLE was 6.43 for squamous cell skin cancer (CI 1.31-18.5, P = 0.012), 5.48 for lymphoma (CI 0.64-19.6, P = 0.052), 2.46 for uterine cancer (CI 0.29-8.78, P = 0.196), 2.0 for ovarian cancer (CI 0.23-7.14, P = 0.264), 1.72 for lung cancer (CI 0.36-4.95, P = 0.254) and 1.6 for breast cancer (CI 0.65-3.23, P = 0.154). The total number of patient-years at risk was 2774 years. The results from this study on an unselected cohort of Icelandic SLE patients do not suggest an overall association between SLE and malignancy. Squamous cell skin cancer was the only individual cancer type that was statistically increased in the population and the numbers for lymphoma were borderline statistically significant.