Introduction: On 1 July 2016, stage 1 of the Tackling Alcohol-Fuelled Violence (TAFV) policy was introduced in Queensland, Australia and included restricted alcohol service to 03:00 in dedicated entertainment precincts (safe night precincts or SNPs). During stage 2 (from 1 February 2017), the number of extended trading permits (i.e. trading until 05:00) per venue were reduced; and during stage 3 (from 1 July 2017), networked identification scanners were mandated for late-night venues. We aim to examine whether patron drinking behaviours in two key SNPs changed significantly following stages 2 and 3. Methods: The study design was repeated cross-sectional, with three data collection (policy) stages. Patrons street surveys were conducted in Fortitude Valley (n = 2066) and Cairns (n = 1021) SNPs between July 2016 and November 2018. Linear or negative binomial regressions were conducted to examine changes in three key outcomes: blood alcohol concentration (BAC) reading, proportion of pre-drinkers and number of pre-drinks. Results: We found no significant difference in BAC reading, proportion of pre-drinkers or number of pre-drinks at stages 2 and 3 of the policy compared to baseline in Fortitude Valley. In Cairns, we found significant reductions in patron BAC at stage 3 of the policy, and female BAC at stages 2 and 3 compared to baseline; while the proportion of females pre-drinking significantly reduced by half in stage 2. Discussion and Conclusions: Despite evidence of some reductions in alcohol consumption, high levels of intoxication remain, suggesting harm reduction may be more difficult to achieve in these drinkers.