The present article reports the factor structure, reliability and validity of Fenigstein et al.'s (1975) Self-Consciousness Scale (SCS) using British respondents from five samples of undergraduate students, two of 17-year-old school pupils, one of 13-year-old pupils, and one of 11-year-old pupils. Confirmatory factor analyses reproduced in most samples a 3-factor structure consistent with the SCS subscales, although exploratory factor analyses suggested the Private subscale consists of several subfactors. Concordantly, alpha reliability coefficients were higher for Social Anxiety and Public than for Private, respectively. Test–retest reliability was reasonable for Social Anxiety, but poorer for Public and Private. In line with correlations found by other researchers, that between Public and Private Self-consciousness was r = 0.33. Both of these subscales correlated positively with Neuroticism. Nevertheless, public and private were considered to be distinguishable constructs. In contrast, Social Anxiety correlated highly positively with Introversion and Neuroticism, and negatively with Self-Esteem, indicating poor discriminant validity. Exploratory factor analyses of scales designed for young adolescents yielded recognizable private and public factors, with the predominance of a Private factor among 11-year olds, but of a Public factor among 13-year olds. Finally, conceptual problems surrounding the private–public distinction are articulated in relation to (a) similar alternative constructs, and (b) Tajfel and Turner's (1979) concept of social identity.