The relationship between behavioral variability and reward expectation has been examined in recent years. This relationship is predictive: when an animal has a low expectation of reinforcement for a particular behavioral set, they engage in high levels of variability in their actions. We conducted two experiments to further investigate this relationship using a novel measure of behavioral variability. In Experiment 1, two groups of rats were trained to travel through a column maze, with many possible reinforced pathways, to receive either their maintenance diet (i.e., chow) or a highly palatable sugary reward (cereal). We hypothesized that animals trained with a maintenance-diet food source (chow) would demonstrate more variation in the pathways taken to the goal location than those animals trained with the highly palatable alternative. In Experiment 2, all rats were trained to travel through the maze to receive alternating outcomes of chow or cereal in a within-subjects design. Results from both experiments indicated that rats emitted greater variability in paths taken to the goal when the reinforcer was the maintenance chow. These results corroborate the relationship between reward and behavioral variability in a new behavioral measure.