The aim of this study was to compare the effectiveness of triamcinolone hexacetonide (THA) and methylprednisolone acetate (MPA), given via the intra-articular route at equipotent dosage to patients with symptomatic knee OA with effusion, in a double-blind randomized comparative trial. Consecutive hospital-referred patients who fulfilled the American College of Rheumatology criteria for knee OA (clinical and radiographic) were randomly allocated to receive either THA 20 mg (1 ml) or MPA 40 mg (1 ml). All patients had synovial fluid aspirated from their knee joint at the time of injection. Assessments were made at 0, 3 and 8 weeks by a second operator, thus blinding both patient and assessor. Outcomes measured at each visit were: knee pain in the previous 48 h (expressed on a 100 mm visual analog scale; VAS), stair climb time (SCT) and Lequesne index score (LEQ). Changes in VAS, SCT and LEQ were compared between the groups using a Student's paired t test. Fifty-seven patients were studied (44 female, 13 male) with a mean age of 62.5 years. Both steroids gave significant pain relief (VAS) at week 3 ( p<0.01) but only MPA showed an effect on VAS and LEQ scores at week 8 compared to baseline ( p<0.05). THA was more effective than MPA at pain reduction at week 3 ( p<0.01); this difference was lost at week 8 ( p=0.17). There was no significant difference between the two drugs in functional endpoints (SCT, LEQ) at either 3 or 8 weeks. Both THA and MPA offer at least temporary symptomatic benefit in knee OA. THA is more effective than MPA at week 3, but its effect is lost by week 8. MPA still has an effect at week 8.