This study investigated the proportional reasoning abilities of 35 college science students. Using a projection of shadows problem, proportional reasoning was assessed under one of two experimental conditions corresponding to the type of information presented: (a) relevant only, or (b) relevant and irrelevant. Subjects were also measured on the cognitive style of field independence. Results showed that subjects in the relevant only condition performed significantly better than those in the relevant-irrelevant condition. Degree of field independence was related to performance only in the relevant-irrelevant condition. These results indicated an interaction between type of information presented in a task and cognitive style.