In an attempt to isolate differential personality characteristics of problem drinkers in New Zealand, 100 problem drinkers, 60 non-problem drinking psychiatric patients, 30 non-problem drinking prison inmates, and 30 abstainers were compared on a number of characteristics extracted from an extensive literature review, preference being given to those which longitudinal studies had indicated might be predisposing to adoption of an addictive rather than a non-addictive pathology. Predictions were confirmed in that problem drinkers manifested diminished investment in the heterosexual role by being covertly orientated to males rather than females (on a measure of seconds spent eye-fixating on pictures of naked males versus females) while verbalizing overtly a preference for the female picture of each opposite-sex pair, from which it was inferred that "defensive masculinity" functioned to obscure covert masculine inadequacy; problem drinkers were socially isolated, with marked difficulties in establishing close relationships and preferring short-term, superficial, casual encounters coupled with withdrawal from stable emotional commitments (biographical data questionnaire); they were also extremely field dependent (embedded figures), intensely aggressive (inventory and stereoscope), manifested a very short future time perspective (personal events methods), and had considerable difficulties in the area of socialization (inventory). They had also experienced home backgrounds in which they were often deprived of sustained contact with a father, either through lengthy separations or the father's emotional withdrawal or antagonistic, rejecting behaviour. The one prediction not confirmed dealt with an attempt to mobilize anxieties over putative dependency and masculine inadequacy, with the expectation that threat to these would result in greater (spurious) self-sufficiency and masculinity scores on inventories, relative to subjects not threatened. While there was some tentative evidence of threat effect on drinkers and abstainers, results were equivocal and the null hypothesis could not be rejected. However, patients differed from the other groups in readily admitting to much less self-sufficiency and masculinity. On all measures drinkers were distinguished from patients (other than in the sharing of social anxieties, though patients were not socially isolated or superficial), and from abstainers (except that drinkers, abstainers, and criminals did not differ in declaring themselves extremely self-sufficient and masculine), but on socialization, and on aggressiveness as measured by the stereoscope, drinkers did not differ from criminals whereas abstainers were hyper-conventional. By comparing young early with old early starters it proved possible to demonstrate that years of chronic inebriety did not influence the characteristics isolated, especially field dependence, aggressiveness, and under-socialization. Younger drinkers were more disturbed than older ones in the areas of aggressiveness and socialization, on inventory measures and on such indices as wanderlust, occupational instability (from preference), impulsiveness, promiscuity, antisocial behaviour during school years (habitual truanting, stealing, disciplinary problems), and having extensive criminal records. 26% whose pathological drinking began after age 40 were much more neurotic-like. Conclusions were that problem drinkers suffered from a personality disorder more severe than a neurosis, and that early socialization difficulties were those most likely to predict later problem drinking.