Objective: To examine behavioural intention to reduce soda consumption after exposure to the Choose Health LA 'Sugar Pack' campaign in Los Angeles County, California, USA. Design: A cross-sectional street-intercept survey was conducted to assess knowledge, attitudes, health behaviours and behavioural intentions after exposure to the 'Sugar Pack' campaign. A multivariable regression analysis was performed to examine the relationships between the amount of soda consumed and self-reported intention to reduce consumption of non-diet soda among adults who saw the campaign. Setting: Three pre-selected Los Angeles County Metro bus shelters and/or rail stops with the highest number of 'Sugar Pack' campaign advertisement placements. Subjects: Riders of the region's Metro buses and railways who were the intended audience of the campaign advertisements. Results: The overall survey response rate was 56 % (resulting n 1041). Almost 60 % of respondents were exposed to the advertisements (619/1041). The multivariable logistic regression analysis suggested that the odds of reporting intention to reduce soda consumption among moderate consumers (1-6 sodas/week) were 1.95 times greater than among heavy consumers (1 soda/d), after controlling for clustering and covariates. Respondents with less than a high-school education and who perceived sugary beverage consumption as harmful also had higher odds; in contrast, respondents aged 65 years had lower odds. Conclusions: Results suggest that future campaigns should be tailored differently for moderate v. heavy consumers of soda. Similar tailoring strategies are likely needed for younger groups, for those with less educational attainment and for those who do not perceive consumption of soda as harmful.