Gambling type involvement, both in terms of participation (engagement in specific gambling types) and diversity (how many gambling types an individual engages in), is a key feature to address in gambling self-report measures, but such systematic measurement procedures are scarce. The aim of this study was to test the psychometric performance of the gambling type assessment in the recently developed Gambling Disorder Identification Test (GDIT), in terms of test-retest reliability, convergent validity, and patterns of gambling diversity, among help-seeking and general population gambling samples (total n = 603). Overall, online gambling was more commonly reported as problematic than land-based gambling. Retest reliability varied for specific gambling types (ICC range 0.32–0.64, r_tet range 0.66–0.85). In terms of gambling participation, online gambling showed stronger correlations with GDIT total score (i.e., symptom severity) than land-based gambling, where Slots showed the strongest correlation (r = 0.52), followed by Casino table games (r = 0.25), Sports and Horse betting (r = 0.16 and r = 0.14, respectively), and Poker (r = 0.14). Lotteries showed no correlation with GDIT total score (r=-0,01). For Slots gambling, all gambling diversity levels (including Slots as a single gambling type) were on average associated with the highest diagnostic severity level (GDIT total score > 30; severe gambling disorder). Finally, explorative configural frequency analysis identified typical and antitypical gambling diversity patterns. The result from the current study corroborates findings that engagement in specific gambling types matter, and that such features should be included in gambling measurement. We conclude that the GDIT is a reliable and valid measure for systematic assessment of gambling type involvement. The GDIT can be used to assess gambling participation and diversity, as part of a broad measurement setup for problem gambling and gambling disorder.