In a context of increasing health problems related to bad dietary habits in developing countries, simplified nutritional label formats may be a promising policy alternative. The aim of this study is to assess whether the multiple traffic light (MTL) system can improve the nutritional quality of consumer decisions. We conducted a selection experiment in Peru, where participants were offered a choice between three alternatives in two different food categories (crackers and beverages). Individuals were randomly assigned products with MTL labels on them. Our results reveal that exposure to MTL labels significantly increases the probability of avoiding the least healthy options and of choosing the healthiest items among the alternatives provided. These effects are large in magnitude and seem to be driven by individuals that are female, report average or above average dietary habits, and lack basic nutritional knowledge. However, our findings also suggest that the effectiveness of MTL labels may be sensitive to the specific characteristics of the set of options provided. These findings contribute to the emerging experimental literature on MTL labels in two ways. First, we provide evidence that this system can be effective in inducing consumers to make healthier choices in real‐life situations. Second, we explore which individuals are more likely to benefit from exposure to MTL labels, as well as potential limitations to their effectiveness. Overall, our results provide new insights on how to assess the issue of bad nutrition in emerging economies.