OBJECTIVE To assess the efficacy of a single intra-articular injection of triamcinolone hexacetonide (THA) in knee osteoarthritis (OA) and examine factors which may relate to treatment efficacy. METHODS Eighty four patients with clinical and radiographic evidence of knee OA were recruited and randomly allocated to receive either THA (20 mg in 1 ml) or placebo (0.9% normal saline, 1 ml). Follow up assessments evaluated the following outcome variables: patient opinion of overall change in the treated knee, visual analogue pain score (VAS), distance walked in one minute (WD), and Health Assessment Questionnaire modified for lower limb function (HAQ). RESULTS Seventy eight percent of THA and 49% of placebo treated patients reported overall improvement at week 1 (p < 0.05). At week 6, improvement was reported in 57% and 55% of patient groups, respectively. VAS improved in both groups at week 1 (THA, p < 0.001; placebo, p < 0.05) and week 6 (both p < 0.01). Improvement in VAS was significantly greater among THA treated patients at week 1 only (p < 0.01). Subgroup analysis of THA treated patients revealed greater improvement in VAS among patients with clinical evidence of an effusion (p < 0.05), and those who had synovial fluid successfully aspirated at the time of injection (p < 0.01). WD improved in THA treated patients at week 1 (p < 0.001), and in both groups at week 6 (THA, p < 0.001; placebo, p < 0.01). Improvements in HAQ were seen in THA patients only at weeks 1 and 6 (p < 0.05). Regression analysis did not identify any additional clinical, radiographic, or synovial fluid characteristics which influenced the response. CONCLUSIONS THA provided short term pain relief in knee OA. Increased benefit was associated with both clinical evidence of joint effusion and successful aspiration of synovial fluid at the time of injection.