In foraging ecology, models of food choice usually assume animals are economically rational, using an absolute valuation system, wherein they choose food with the highest net worth and ignoring irrelevant alternatives (e.g., lower value items). However, studies have found that animal decisions can be influenced by the specific composition of a choice set, suggesting they use comparative valuation systems. The prevalence of comparative decisions could be driven by nutrition focused decision strategies, or how and when we measure preference. Herbivores are an important group of organisms, yet their valuation system has not been explicitly tested. I aimed to explore different aspects of rationality with herbivores. First, I tested if the timeframe in which preference is measured influences whether free ranging brushtail possums are rational. I measured possum preferences for synthetic foods over three timeframes in the presence or absence of a phantom decoy (high value but unavailable item). I found that the decoy changed possum preference in the shortest timeframe, but not in the longer timeframes. In chapter 3, I explored the drivers of preference with captive herbivores. I used synthetic foods to test if the definition of preference or nutritional content influenced snail preference. I found that changing the definition of preference influenced how many snails were transitive. I was unable to find a nutritional explanation for snail food preferences. Finally, I tested whether snails were rational when presented with different vegetables. I found that snails respected transitivity, and while there was no decoy effect at the group level, individuals that showed a preference for the least preferred option in the binary treatment were more likely to change their preference in the presence of a decoy. My work suggests that part of the difficulty with rationality research is that whether a principle is respected can be dependent on how and when preference is measured.