This study concerns the prevalence of loneliness in a sample of adolescent Australian college students and examines the predictors of loneliness in this group, in particular the impact of social network characteristics, social network appraisal, the functions of friendship, and the psychosocial variables of identity and intimacy (Erikson, 1978) within the context of de Jong-Giervald's (1987) model of loneliness. One hundred and thirty- eight college students aged between 17 and 20 years from one regional and one city tertiary institution were surveyed to ascertain the level of experienced loneliness as measured by the UCLA Loneliness Scale (Russellet al., 1980). Measures of intimacy and identity were obtained using the Erikson Psychosocial State Inventory (Rosenthalet al., 1981) while the functions of friendship in general and of a specific same- sex close friend were those established in a previous study (Moore and Boldero, 1987). Comparison of the levels of loneliness reported with those found by other studies suggested that Australian adolescents are no less vulnerable to the experience of loneliness than their American counterparts. In addition, while no quantitative sex differences in loneliness were found, qualitative differences emerged. Specifically, although psychosocial intimacy was the best predictor of loneliness for both male and female, the other independent predictors were different. For males having fewer same- sex friends and an inability to share feelings with a specific same- sex friend predicted loneliness whereas for females social network appraisal, the ability to engage in mutual aid with a specific same-sex friend, living with family, and psychosocial identity were important. These differences were discussed in terms of the impact that the differential socialization of male and female has on ability of adolescents to engage in close and satisfying interpersonal relationships. Suggestions for intervention were made.