Objective While most research on minority representation in Congress finds that the African‐American constituency size influences representation of the group's interests, most recent studies examining Latinos find no such relationship. We argue that the failure to find a relationship stems from the focus of prior research on the proportion of the Latino population in the total geographic constituency rather than the proportion of Latinos in the electoral constituency, what we term “Latino electoral strength” (LES). Methods Using data on Latino turnout at the state level, we examine the effect of LES on representation of Latino interests in the U.S. Senate. We use ordinary least squares (OLS) regression that accounts for the time serial, cross‐sectional (TSCS) nature of our data. Results Consistent with other studies, we find no influence of Latino population size on Latino representation. However, LES has a significant, negative effect on Latino representation. Conclusions Our results suggest that greater LES leads to worse representation of their interests, and that this relationship increases as LES grows. This is consistent with existing studies of racial threat theory.