To perform a systemic review and meta-analysis on the effectiveness of radiosynoviorthesis (RSO). A search of medical databases was conducted. Criteria for inclusion: articles in English, minimum follow-up of 6 months, specification of joint disease, reported outcome of at least 5 RSOs. The studies were scored for quality by the Oxford Centre of Evidenced-based Medicine Levels of Evidence, from 1 to 4. Twenty-one (21) studies were included (3 quality 1b, 5 2b and 13 4), analysing 169Erbium/186Rhenium-RSO used predominantly in small joints and 49 (1 quality 1b, 10 2b and 38 4) on 90Yttrium-RSO used predominantly in knee joints. The reported success rates of 169Erbium/186Rhenium-RSO ranged from 69-100% at 6 months, and from 54-100% at > or =12 months; for 90Yttrium they were 24-100% and 29-94%, res-pectively. Studies comparing the effect of RSO with that of glucocorticoid (GC) or saline injection alone were pooled. At 6 months, the pooled odds ratio favouring RSO of the knee with Yttrium over control is 4 (confidence interval (CI) 95% 1.2-14), p=0.02, but at 12 months the ratio was 1.7 (CI95% 0.69-4), p=0.26. For RSO of small joints with Erbium/Rhenium compared to controls, the pooled odds ratio at 6 months is 2 (CI95% 0.66-6), p=0.22 and at 12 months 2 (CI95% 1.09-3.5), p=0.03. Reported success rates of RSO are high, but differences in effect with GC injection are less evident, although there is marked heterogeneity in study design of the (small number of) comparative studies.