Ninety-five women undergoing induced abortion were randomly premedicated with oral diazepam, 5 or 10 mg, midazolam 15 mg, or intramuscular placebo, 40-60 min before the induction of anesthesia. Prior to premedication and again prior to the procedure, the women completed the questionnaire sheet for the Profile of Mood States (POMS), and plasma samples for immunoreactive beta-endorphin (ir beta-E) and ACTH were taken. The Oblique Tension-Anxiety (T-A) Factor scores derived from POMS correlated with plasma concentrations of ir beta-E and ACTH after premedication. The T-A scores decreased in women premedicated with diazepam, 5 and 10 mg, or midazolam, 15 mg, but not in women treated with placebo. The women premedicated with midazolam, 15 mg, became more fatigued after premedication. The changes in blood pressure after premedication correlated with T-A scores. A positive correlation was found between heart rate and plasma beta-endorphin concentration. The changes in ir beta-E and ACTH did not correlate with the changes in T-A scores. We conclude that POMS T-A scores are useful for assessment of preoperative anxiety and the effect of premedication. The present study did not provide any reliable proof to confirm the hypothesis of a relationship between plasma concentrations of ir beta-E or ACTH and preoperative anxiety. Since many factors modulate endorphin and ACTH secretion prior to operation, the measurement of endogenous opiates may be of limited value in assessment of the effects of preanesthetic medication.