Aims: For the purpose of analyzing the relationship between having participated in particular gambling activities, the overall gambling involvement and the occurrence of associated gambling problems, data of three representative surveys on gambling behaviour in the population aged 16 to 65 in Germany 2007, 2009 and 2011 were matched. Methods: Telephone sampling (CATI) surveying the most relevant gambling activities, standardized instruments: South Oaks Gambling Screen (SOGS) for screening on gambling addiction, bi- und multivariate logistic regression analyses, dependent variable: problem gambling behaviour (problematic or pathological classification by SOGS). Results: Compared to recreational gamblers most gambling activities in the preceding 12 months were stated far more often by problem gamblers. Even more the latter report a higher overall gambling involvement (number of different gambling activities). Highest proportions of problem gamblers are to be found among those respondents, who used internet casino games, electronic gaming machines (EGMs), sports betting from private providers, casino games, and the lottery 'Keno' (percentages in descending order: 11,7 % to 6,5 %). Accordingly bivariate regression analyses reveal for these gambling activities in roughly the same order the highest risks for problem gambling. Though lowering down considerably when additionally controlling for overall gambling involvement the risks of the mentioned gambling activities with the exception of casino table games remain significant. Looking at all gambling activities simultaneously in multivariate analyses, thereby taking into account the individual combination of gambling activities, reveal highest risks for EGMs in gambling halls, pubs etc., followed by internet casino games (Odds Ratios 4.77 resp. 3.75). Independent of statistical modelling, the risks for lotteries are comparatively low. Conclusions: If analyzing relations between the occurrence of problem gambling and the use of distinct gambling types, by reason of frequent multiple gambling activities of problem gamblers, the overall gambling involvement should also be taken into account. Following the results regulation and prevention actions should predominantly address gambling at EGMs in gambling halls, pubs etc., and internet casino games.