An emerging literature has identified optimal low-risk gambling limits in an effort to reduce gambling-related harm. Concerns have, however, been raised about the construction of aggregate low-risk limits that are applied to all gambling activities and there is support from gambling experts and the general public in Australia for the identification of low-risk limits for specific gambling activities. The study's aim was to identify and evaluate a set of empirically-based activity-specific limits (gambling frequency, gambling expenditure, gambling expenditure as a proportion of gross personal income, session expenditure, session duration) in a secondary analysis of Social and Economic Impact Studies of Gambling in Tasmania and the 2014 Survey on Gambling, Health and Wellbeing in the ACT. Balancing sensitivity and specificity, limits were identified for all gambling activities: EGMs (10 times per year, AUD$300/year, 0.63-1.04% of personal income, AUD$35 per session, 40 min/session), horse/dog racing (0.55% of personal income), instant scratch tickets (AUD$45/year), lotteries (0.45% of personal income), keno (4-13 times/year, AUD$45-$160/year), casino table games (AUD$345/year, 0.36-0.76% of personal income), bingo (AUD$150/year, 0.49% of personal income, AUD$17/session, 90 min/session), and sports/other event betting (14 times/year, AUD$400/year, 0.55-0.86% of personal income). These limits were exceeded by one-quarter to one-half of gamblers on these specific activities and were generally good predictors of gambling-related harm in subgroups of gamblers participating in these gambling activities and in the overall gambling sample. The limits provide gamblers, regulators, prevention workers, and researchers with simple rules of thumb in prevention efforts to reduce gambling-related harm in specific contexts.