We investigated the population-level relationship between exposure to brand-specific advertising and brand-specific alcohol use among US youth. We conducted an internet survey of a national sample of 1031 youth, ages 13-20, who had consumed alcohol in the past 30 days. We ascertained all of the alcohol brands respondents consumed in the past 30 days, as well as which of 20 popular television shows they had viewed during that time period. Using a negative binomial regression model, we examined the relationship between aggregated brand-specific exposure to alcohol advertising on the 20 television shows [ad stock, measured in gross rating points (GRPs)] and youth brand-consumption prevalence, while controlling for the average price and overall market share of each brand. Brands with advertising exposure on the 20 television shows had a consumption prevalence about four times higher than brands not advertising on those shows. Brand-level advertising elasticity of demand varied by exposure level, with higher elasticity in the lower exposure range. The estimated advertising elasticity of 0.63 in the lower exposure range indicates that for each 1% increase in advertising exposure, a brand's youth consumption prevalence increases by 0.63%. At the population level, underage youths' exposure to brand-specific advertising was a significant predictor of the consumption prevalence of that brand, independent of each brand's price and overall market share. The non-linearity of the observed relationship suggests that youth advertising exposure may need to be lowered substantially in order to decrease consumption of the most heavily advertised brands.