Alcohol screening and brief intervention for unhealthy alcohol use has not been consistently delivered in primary care as part of preventive healthcare. To explore whether telephone-based intervention delivered by a health educator is efficacious in reducing at-risk drinking among older adults in primary care settings. Secondary analyses of data from a randomized controlled trial. Subjects randomized to the intervention arm of the trial (n = 310). Personalized risk reports, advice from physicians, booklet about alcohol and aging, and up to three telephone calls from a health educator. All interventions were completed before the three-month follow-up. Risk outcomes (at-risk or not at-risk) at 3 and 12 months after enrollment. In univariate analyses, compared to those who remained at risk, those who achieved not at-risk outcome at 3 months were more likely to be women, Hispanic or non-white, have lower levels of education, consume less alcohol, drink less frequently, and have lower baseline number of risks. In mixed-effects logistic regression models, completing all three health educator calls increased the odds of achieving not at-risk outcome compared to not completing any calls at 3 months (OR 5.31; 95% CI 1.92–14.7; p = 0.001), but not at 12 months (OR 2.01; 95% CI 0.71–5.67; p = 0.18). Telephone-based intervention delivered by a health educator was moderately efficacious in reducing at-risk drinking at 3 months after enrollment among older adults receiving a multi-faceted intervention in primary care settings; however, the effect was not sustained at 12 months.