In order to assess the analgesia obtained from single oral doses of paracetamol alone and in combination with codeine in postoperative pain, we conducted a systematic review of randomised controlled trials. We found 31 trials of paracetamol against placebo with 2515 patients, 19 trials of paracetamol plus codeine against placebo with 1204 patients and 13 trials of paracetamol plus codeine against the same dose of paracetamol with 874 patients. Pain relief information was extracted, and converted into dichotomous information (number of patients with at least 50% pain relief). Wide variations in responses to placebo (0-72%) and active drug (3-89%) were observed. In postoperative pain states paracetamol 1000 mg alone against placebo had an number-needed-to-treat (NNT) of 3.6 (3.0-4.4) and paracetamol 600/650 mg alone an NNT of 5.0 (4.1-6.9). Paracetamol 600/650 mg plus codeine 60 mg against placebo had a better NNT of 3.1 (2.6-3.8), with no overlap of 95% confidence intervals with paracetamol 600/650 mg alone. In direct comparisons of paracetamol plus codeine with paracetamol alone the additional analgesic effect of 60 mg of codeine added to paracetamol was 12 extra patients in every 100 achieving at least 50% pain relief. In indirect comparisons of each with placebo it was 14 extra patients per 100. This was an NNT for adding codeine 60 mg of 9.1 (5.8-24). The results confirm that paracetamol is an effective analgesic, and that codeine 60 mg added to paracetamol produces worthwhile additional pain relief even in single oral doses.