Psychologic and demographic characteristics have been shown to correlate with behavioral compliance in studies of prevalent hemodialysis (HD) patients. Few data, however, exist on the psychologic characteristics or their relationship to compliance in patients initiating HD treatment, or incident patients. Our previous work with prevalent HD patients showed an interrelationship among social support, perception of the effects of illness, and depression measured by a standardized depression index and a cognitive depression index deleting somatic symptoms. To compare psychosocial and behavioral compliance parameters in incident and prevalent patients, 99 incident HD patients undergoing dialysis for less than 6 months were assessed by using a variety of questionnaires. The patients' mean standardized depression index score was in the range of mild depression. There was no difference in mean age, Karnofsky functional status, perception of effects of illness or standardized or cognitive depression scores between incident patients and 149 prevalent patients treated with HD for 6 to 246 months. Satisfaction with life scores and perception of sexual adjustment were better, but Kt/V (quantitative measure of the amount of dialysis provided) and protein catabolic rate were lower in incident than in prevalent patients. Social support and perception of illness scores correlated with depression measures in incident patients, findings that were similar to prevalent patients. Incident (as did prevalent) patients showed striking bivariate correlations between the psychosocial and depression variables, but fewer correlations between psychologic and compliance measures. Both standard compliance parameters, such as mean predialysis serum phosphorus and potassium concentration, and behavioral compliance were better in incident than in prevalent patients. Worsened sexual adjustment, functional status, and increased severity of illness were correlated with improved behavioral compliance in incident patients. Behavioral compliance styles correlated with different social support and severity of medical illness variables in incident and prevalent patients, when assessed by multiple linear regression analysis. These different patterns suggest the existence of different mechanisms of adaptation in the groups. Measures focused on increasing delivery of dialysis and enhancing social support in urban African-American patients starting HD therapy may be useful in improving compliance with the dialysis prescription, and indirectly improving survival.