AIMS: This study seeks to establish the prevalence alcohol-related harms to children (ARHC) that occur because of others' drinking in the general population and examine how this varies by who was reported to have harmed the child and socio-demographic factors. DESIGN AND SETTING: A randomly selected cross-sectional national population telephone survey undertaken in 2008 in Australia. PARTICIPANTS: A total of 1142 adult respondents who indicated they lived with or had a parental/carer role for children. MEASUREMENTS: Questions included whether children had been negatively affected in any way, left unsupervised or in an unsafe situation, verbally abused, physically hurt or exposed to serious family violence because of others' drinking in the past year. FINDINGS: Twenty-two per cent of respondents reported children had been affected because of another's drinking in the past year; 3% reported substantial harm. Respondents most commonly reported that children were verbally abused because of others' drinking (9%). Participants in single-carer households were more likely to report ARHC than participants in households with two carers, and participants who drank weekly were more likely to report ARHC than those who did not drink. CONCLUSIONS: Almost a quarter of those with a caring role for children in Australia reported that a child or children with whom they lived or for whom they were responsible have been affected adversely by others' alcohol consumption in the past year. The problem extends across the social spectrum, but children in single-parent homes may be at higher risk.