To evaluate the effect of licofelone, an arachidonic acid substrate with combined inhibitory activity against 5-lipoxygenase and cyclooxygenases 1 and 2, a double-blind, randomised and placebo-controlled study was conducted in 33 client-owned dogs that were lame owing to hindlimb osteoarthritis. Seventeen of the dogs received a placebo and 16 were treated with 2·5 mg/kg licofelone twice a day for 28 days. The dogs' lameness was assessed on a visual analogue scale (vas), and by force plate analyses at baseline and 14 and 28 days after starting the treatment. After 14 days the mean (se) change in peak vertical force in the licofelone-treated dogs (1·7 [0·8] per cent bodyweight) was significantly greater (P<0·05) than in the placebo-treated dogs (−0·3 [0·6] per cent bodyweight), and after 28 days the difference had increased. In contrast, the dogs' lameness, as assessed by the vas values, had decreased significantly over baseline in both the treated and control groups.