BACKGROUND Hydroxychloroquine is used for the treatment of rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE). Long term studies have shown a high rate of termination of hydroxychloroquine treatment in patients with RA. Although it has been shown that discontinuation of treatment with hydroxychloroquine is associated with exacerbation of SLE, long term maintenance rates of treatment with hydroxychloroquine in patients with SLE have not been investigated. METHODS Hydroxychloroquine use in patients with RA and SLE in a group of patients in a single community rheumatology practice was studied. Information was drawn from a computer drug use database containing details of the beginning and end of treatment. Data were analysed using life table methods. RESULTS Four hundred and three treatment episodes (366 patients with RA, 37 patients with SLE) were observed over eight years. In patients with RA, the cumulative probability of discontinuing treatment was 37% at 12 months and 54% at 24 months. In contrast, hydroxychloroquine treatment of patients with SLE continued over significantly longer periods of time (p < 0.001); the discontinuation probabilities at 12 and 24 months were 8 and 24% respectively. Treatment terminations were predominantly for inefficacy; terminations for toxicity were limited to the first 19 months of treatment. No ocular toxicity was observed. CONCLUSIONS Treatment of patients with RA in a community rheumatology practice with hydroxychloroquine has a low probability of long term continuation, mostly because of inadequate control of disease manifestations rather than toxicity. In patients with SLE, treatment with hydroxychloroquine has a significantly higher probability of long term continuation.