Many social cognitive assessment measures that are appropriate for clinical use are currently available, but there is a general concern about their ecological validity. This study aimed to develop an applicable real interaction-based test to assess social cognition. A sample of 50 subjects (mean age 22 ± 5.8, 56% women) took the Social Interaction Test as well as two instruments for assessing social cognition: (1) the Movie for Assessment of Social Cognition (MASC) and (2) branch 4 from the Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT). The test showed no incidence on its application. The reliability of the 18-item final version of the test was a medium-high level (Cronbach’s alpha = 0.701). To assess the internal structure of the test, a multidimensional scaling procedure was used. The common space of coordinates for the two-dimensional solution showed a normalized raw stress of 0.076 and Tucker’s congruence coefficient of 0.965. The social interaction test showed stronger association with MASC (more realistic, video-based format) than with MSCEIT (less realistic, paper-based format). The Social Interaction Test is applicable and feasible to use it to assess social cognition in the general population.