A double-blind randomized trial was carried out in 90 male patients suffering from pain after meniscectomy. The patients received a single dose of paracetamol 1000 mg plus codeine 60 mg, paracetamol 1000 mg, codeine 60 mg, or placebo. The tablets were taken when needed after surgery and the postoperative pain was recorded on a visual analogue scale. Over a period of 4 h the efficacy of the drugs was calculated in terms of pain intensity, pain intensity difference and percentage pain reduction. The greatest effect was obtained in patients taking the paracetamol plus codeine combination. Statistical analysis was carried out by use both of parametric and non-parametric procedures. The results suggest that pain reduction is a valuable measurement of analgesic efficacy and that non-parametric assumptions are preferable in the statistical analysis of analgesic activity.