BACKGROUND AND AIMS:To reduce health and social inequities, it is important to understand how drinking patterns vary within and between Indigenous peoples. We aimed to assess variability in estimates of Indigenous Australian drinking patterns and to identify demographic and methodological factors associated with this. DESIGN:A three-level meta-analysis of Australian Aboriginal and Torres Strait Islander ('Indigenous') drinking patterns [International Prospective Register of Systematic Reviews (PROSPERO) no. CRD42018103209]. SETTING:Australia. PARTICIPANTS:Indigenous Australians. MEASUREMENTS:The primary outcomes extracted were drinking status, single-occasion risk and life-time risk. Moderation analysis was performed to identify potential sources of heterogeneity. Moderators included gender, age, socio-economic status, local alcohol restrictions, sample population, remoteness, Australian state or territory, publication year, Indigenous involvement in survey design or delivery and cultural adaptations. FINDINGS:A systematic review of the literature revealed 41 eligible studies. For all primary outcomes, considerable heterogeneity was identified within ( I22 = 51.39-68.80%) and between ( I32 = 29.27-47.36%) samples. The pooled proportions (P) of current drinkers [P = 0.59, 95% confidence interval (CI) = 0.53-0.65], single-occasion (P = 0.34, 95% CI = 0.24-0.44) and life-time (P = 0.21, 95% CI = 0.15-0.29) risk were all moderated by gender, age, remoteness and measurement tool. Reference period moderated proportions of participants at single-occasion risk. CONCLUSIONS:Indigenous Australian drinking patterns vary within and between communities. Initiatives to reduce high-risk drinking should take account of this variability.