Fifty-nine patients commencing intramuscular sodium aurothiomalate therapy were randomized to receive three doses of 120 mg intramuscular depot methylprednisolone acetate or matching placebo at weeks 0, 4, and 8 in addition to chrysotherapy. The group receiving methylprednisolone had more rapid disease improvement. This advantage persisted for up to 12 weeks, although by 24 weeks both groups exhibited similar benefits due to continued improvement in the group treated with gold alone. Withdrawals secondary to gold-induced side-effects occurred later in the steroid group (median time to withdrawal: 15 weeks steroid; 4.5 weeks placebo, P less than 0.05), and there were fewer withdrawals due to a lack of effect in the steroid group (one steroid versus three placebo, P NS). We conclude that glucocorticoids given as intermittent, intramuscular depot injections have a significant short term benefit which can be maintained by concomitant administration of intramuscular gold.