Objective: To compare acupuncture with sham (placebo) acupuncture for treatment of nausea of pregnancy. Methods: In a subject- and observer-masked, randomized, controlled trial in the maternity unit at Exeter Hospital, we gave 55 women between 6 and 10 weeks’ gestation genuine, traditional-style acupuncture or sham treatment with a cocktail stick on three or four occasions over 3 weeks. The main outcome measure was nausea score, as determined by subject report on a visual analogue scale in a daily diary. Anxiety and depression also were assessed. Results: Nausea scores decreased from a median of 85.5 (interquartile range 71.25–89.75) to 47.5 (interquartile range 29.25–69.5) in the acupuncture group and from 87.0 (interquartile range 73.0–90.0) to 48.0 (interquartile range 14.0–80.0) in the sham treatment group. There was strong evidence of a time effect (P < .001) but no evidence of a group effect (P = .9) or a group-time interaction (P = .8). Similarly, there was evidence of time effects in scores for anxiety and depression but no group differences. The study had a power of 95% to detect significant differences in nausea scores. Conclusion: Acupuncture was as effective in treating nausea of pregnancy as a sham procedure.