We aimed to test the effects of providing municipal support and training to primary health care providers compared to both training alone and to care as usual on the proportion of adult patients having their alcohol consumption measured.We undertook a quasi-experimental study reporting on a 5-month implementation period in 58 primary health care centres from municipal areas within Bogotá (Colombia), Mexico City (Mexico), and Lima (Peru). Within the municipal areas, units were randomized to four arms: (1) care as usual (control); (2) training alone; (3) training and municipal support, designed specifically for the study, using a less intensive clinical and training package; and (4) training and municipal support, designed specifically for the study, using a more intense clinical and training package. The primary outcome was the cumulative proportion of consulting adult patients out of the population registered within the centre whose alcohol consumption was measured (coverage).The combination of municipal support and training did not result in higher coverage than training alone (incidence rate ratio (IRR) = 1.0, 95% CI = 0.6 to 0.8). Training alone resulted in higher coverage than no training (IRR = 9.8, 95% CI = 4.1 to 24.7). Coverage did not differ by intensity of the clinical and training package (coefficient = 0.8, 95% CI 0.4 to 1.5).Training of providers is key to increasing coverage of alcohol measurement amongst primary health care patients. Although municipal support provided no added value, it is too early to conclude this finding, since full implementation was shortened due to COVID-19 restrictions.Clinical Trials.gov ID: NCT03524599; Registered 15 May 2018; https://clinicaltrials.gov/ct2/show/NCT03524599.