The present study was carried out to construct and validate a self-report measure of alexithymia (AS - 21). After pilot study, the factorial validity and internal consistency of the scale was determined on a sample of 863 subjects (men = 429, women = 434) of different age groups recruited from different cities of Punjab. Thirtyeight items were subjected to principal component analysis through Varimax rotation method. Results revealed a three factor solution with .82 alpha reliability. The three factors were named as Difficulty in recognizing emotions, Difficulty in expressing feelings and Lack of introspective awareness. Factor solution obtained by EFA was further confirmed and validated by the confirmatory factor analysis. Furthermore results revealed significant gender differences in alexithymia. Convergent validity of AS-21 was proved by finding positive correlation with Toronto Alexithymia Scale. The Alexithymia scale is a promising measure with good items homogeneity, internal consistency and a meaningful pattern of validity.