Models of long-term treatment in recurrent unipolar illness ideally should integrate both psychological and biological factors. In earlier reports we noted that high treatment specificity (i.e. good-quality maintenance interpersonal psychotherapy) and high delta sleep ratio were each associated with significantly increased wellness intervals in the absence of pharmacotherapy among patients with recurrent unipolar depression. To determine how these specific factors when taken together are related to length of survival time, we examined the concurrent effects of treatment specificity and delta sleep ratio on wellness intervals using survival analysis. We found significant effects of both treatment specificity and delta ratio on survival time. Seventy-three per cent of the patients in the high treatment specificity/high delta ratio group survived the 3-year trial, while 44% of the patients in the low delta ratio but high treatment specificity group survived. None of those rated low on both variables survived. We also found an effect for individual clinicians on treatment specificity and survival time and noted that the prophylactic effect of treatment specificity was maintained even within subsets of therapists grouped by their patients' survival times. Secondary analyses revealed an effect of patient attitudes on treatment specificity and survival time, although, when taken together, treatment specificity was the only variable remaining significantly associated with outcome. We conclude that patients remain well the longest when pre-treatment delta sleep parameters more closely approximate those of non-depressed individuals and when monthly psychotherapy is of higher quality. The key finding is that high specificity is of significant prophylactic benefit even for patients with a biological vulnerability for recurrence. We also conclude that in addition to therapists, patient expectancies contribute to treatment specificity, and high treatment specificity is, in turn, reflected in longer times to recurrence.