Introduction/objectives: The DSM-IV indicates severity of social phobia (SP) by the “generalized subtype”, when “most social situations” are feared. This specifier refers to the number of feared social situations, perhaps ignoring quantitative differences. We therefore compared specific and interaction-related vs. performance-related social fears according to clinical (age of onset, avoidance, impairment, comorbidities) and vulnerability characteristics (behavioural inhibition (BI), parental psychopathology and rearing). Methods: Six social situations and SP along with their clinical characteristics were assessed using the Munich-Composite International Diagnostic Interview (DIA-X/M-CIDI) in a population-based sample of N=3,021 14-24 year olds. BI and parental rearing were assessed using self-report questionnaires. Parental psychopathology was assessed directly in parents via DIA-X/M-CIDI, supplemented by offsprings’ family-history reports. Results: Isolated social fears were rare, except for fear of taking tests and public speaking. The majority reported to fear two or more social situations. Compared to isolated fears of either interaction or performance situations, their co-occurrence was associated with lower age of onset, severe avoidance and impairment, more comorbid anxiety and depressive disorders. All social fears (in particular interaction-related fears) were associated with higher BI. Associations with parental psychopathology and unfavourable rearing were less consistent, albeit strongest for interaction-related fears. Conclusions: Interaction-related fears may represent a more familial form of SP, while performance-related fears might be less impairing and originate from non-familial factors. The DSM-IV specifier of SP may overlook these differences when individuals with predominantly interaction-related fears are categorized as generalized SP-cases. Findings suggest considering alternative specifiers for SP in future diagnostic systems. [Copyright &y& Elsevier]