This retrospective study determined the prevalence and nature of hospitalisations secondary to infection, and examined the mortality from infection in our large British cohort of patients with systemic lupus erythematosus (SLE). Casenote and database information of 104 consecutive patients attending the UCLH specialised SLE clinic were reviewed for the number of hospitalisations due to infection and the clinical and serological features of affected patients. Cohort mortality data were examined to identify deaths secondary to infectious diseases. Infection serious enough to result in hospitalisation occurred in 15% of the patients in the selected sample of our whole cohort. Six patients had more than one admission due to infection, with pneumonia being the most frequent. Typical bacterial pathogens were most commonly identified. A significant association between admissions for infection and nephritis (P < 0.05 by Chi-square) was identified; however, the use of high dose prednisolone or other immunosuppressants did not increase the risk for infection requiring hospitalisation (P > 0.5 by Chi-square) in our study. Between 1978 and 2007, 17 of 67 (25%) deaths in our SLE cohort of 470 patients were because of infection. Patients who died from infectious causes were more likely to have existing or previous renal disease than those who died from non-infectious causes (P < 0.01 by Chi-square). The majority who died from infection were on high dose prednisolone plus at least one other immunosuppressive agent and had serologically active disease. The study highlights the significant problem of infection in British patients with SLE. Early recognition and treatment of infectious diseases in these patients together with considered use of immunosuppressant medications and vaccinations may help to reduce the impact of these complications.