Fifty-nine women completed a randomized, double-blind placebo-controlled study of pyridoxine hydrochloride (vitamin B6) for the treatment of nausea and vomiting of pregnancy. Thirty-one patients received vitamin B6, 25-mg tablets orally every 8 hours for 72 hours, and 28 patients received placebo in the same regimen. Patients were categorized according to the presence of vomiting: severe nausea (score greater than 7) or mild to moderate nausea (score of 7 or less). The severity of nausea (as graded on a visual analogue scale of 1-10 cm) and the number of patients with vomiting over a 72-hour period were used to evaluate response to therapy. Twelve of 31 patients in the vitamin B6 group had a pre-treatment nausea score greater than 7 (severe) (mean 8.2 +/- 0.8), as did ten of 28 patients in the placebo group (mean 8.7 +/- 0.9) (not significant). Following therapy, there was a significant difference in the mean "difference in nausea" score (ie, baseline - post-therapy nausea) between patients with severe nausea receiving vitamin B6 (mean 4.3 +/- 2.1) and placebo (mean 1.8 +/- 2.2) (P less than .01). In patients with mild to moderate nausea and in the group as a whole, no significant difference between treatment and placebo was observed. Fifteen of 31 vitamin B6-treated patients had vomiting before therapy, compared with ten of 28 in the placebo group (not significant). At the completion of 3 days of therapy, only eight of 31 patients in the vitamin B6 group had any vomiting, compared with 15 of 28 patients in the placebo group (P less than .05).(ABSTRACT TRUNCATED AT 250 WORDS)