Proportional reasoning is essential for many real-world tasks, yet its developmental trajectory remains debated. Children’s performance in nonsymbolic proportional reasoning varies across tasks and plummets when numerical information is misleading. The present study investigates whether 5- to 7-year-old children can accurately compare proportions in a naturalistic context where counting strategies are ineffective. Children listened to short stories in which a subset of people from each of two groups experienced an event (e.g., catching the flu). Given the equal numbers of affected individuals in both groups and different group sizes, children needed to rely on proportional reasoning to compare the prevalence of the event. Results showed that children performed significantly above chance overall. Moreover, they were more accurate in adverse scenarios (e.g., avoiding illness) than in favorable ones (e.g., acquiring rewards). These preliminary findings suggest that the ability to compare nonsymbolic proportions emerges by age 5 but varies depending on context.