The researchers developed low-risk limits for eight gambling activities using data from two large population-based studies in Australia. The limits were based on five indices: (1) gambling frequency; (2) gambling spending per year; (3) spending as a proportion of gross personal income; (4) spending per session; and (5) duration per session. The researchers were able to identify low-risk limits across all five indices for electronic gaming machines (EGMs) only. The limits for EGMs were generally strongest at predicting harm. This was followed by the limits for casino table games and sports/other event betting. The limits for instant scratch tickets, lottery, and keno could also predict harm. However, the limits for bingo and horse/dog racing failed to predict harm.