Over a decade of research since the publication of Human Sexual Inadequacy by Masters and Johnson has confirmed the effectiveness of psychological treatment for sexual problems, but has not advanced our understanding of aetiology very much. It is generally assumed that anxiety blocks normal function in sexual disorders such as impotence or frigidity, although hormonal changes have also been suggested. Evidence for either proposition is lacking. Our own attempts to understand mechanisms have included the use of a dismantling strategy to investigate the effective components of psychological treatment, and latterly the role of exogenous androgens. In our first study, brief treatment modelled on that of Masters and Johnson proved superior to another version of graded practice in sexual contact which lacked the counselling component. Similarly, the fuller version had more effect than did imaginal desensitization, suggesting that sexual dysfunctions cannot be treated like phobic anxiety. A second study focused on women complaining of little sexual interest or enjoyment. The use of a minor tranquillizer was compared with that of androgen (testoral) in the hope that different treatments would be shown to be best for different types of disorder. Unexpectedly, the androgen proved best overall while another variable, monthly vs weekly sessions, did not produce any differences in effect. In the most recent study, a further 48 women were given androgens or placebo, and were seen weekly or monthly for four months by either one or two therapists. There were few significant differences, the most consistent being greater subjective improvement by the women being seen at weekly intervals. The failure to find a positive hormone effect in comparison with placebo raises the possibility that the anxiolytic used earlier may have been counter-productive. If so, it seems unlikely that either excessive anxiety or simple androgen deficiency is an adequate aetiological explanation for female sexual dysfunction.