The dimensionality of the UCLA Loneliness Scale was assessed in 326 adolescents aged 12 to 14. Based on theoretical grounds, a one-factor solution was examined. The results lend limited support to the assertion that the instrument measures a single dimension of loneliness. Factor solutions ranging from three to five were unacceptable according to the methodological criteria used in this study. A two-factor solution met the methodological criteria but the results were suggestive of a method artifact. This possibility was explored using a procedure that follows the logic of construct validation. The findings suggested that the instrument has two substantive dimensions. Continued support of the bidimensionality of the instrument would have implications for theory development regarding the phenomenon of loneliness.