Recently disordered gambling is reclassified as an addictive disorder which inter alia affects a little but significant proportion of adolescents. The aim of this study is to identify and assess different levels of gambling severity among adolescent gamblers (N=1157) from middle and high schools of Korca region, utilizing a cross-sectional design and self-report questionnaire. Using Exploratory, Parallel, Reliability and Confirmatory Analysis, the PGSI measurement instrument, designed for these purposes, resulted to be an appropriate unidimensional screening tool of adolescent disordered gambling in terms of psychometric properties. A traditional Latent Class Analysis using the nine PGSI-items as indicators is performed to identify and predict subtypes of adolescent gamblers, classifying them into latent classes based on their problem gambling severity levels. The inclusion of three covariates related to adolescent gambling problems into Latent Class Model improved the model, helping us to better understand the latent structure.