A number of studies have looked at whether the percentage of visible eye whites could be a reliable and dynamic tool for measuring emotional state in cattle. In this study we have built upon previous research to further test this measure with different stimuli and different types of emotional states in order to assess its suitability as a welfare tool. We used positive and negative contrasts to elicit the emotional states of excitement and frustration in 22 Holstein dairy cows. We performed 10, 15 min focal observations with each cow. In the first four trials the cows were given standard feed, a substrate they have continuous access to. Then for the next five trials they were given concentrates, a high energy feed that is highly desired, and which they have limited access to. And for the final trial they were given inedible woodchip. The standard feed represented a neutral stimulus as it wasn’t novel or highly desirable. The concentrates were a positive stimulus, and the inedible woodchip was a negative stimulus, especially as it followed the concentrates, and so the cow’s expectations were thwarted. We measured both the cow’s heart rate (beats per minute), and the percentage of visible eye whites throughout the focal observations. We found that the woodchip treatment elicited the highest heart: pre-feeding, M = 83.01 feeding, M = 88.95 and post-feeding M = 84.51, suggesting it was the most arousing of the three treatments, this was followed by the concentrates treatment. Results showed that the percentage of visible eye white significantly increased during the concentrates and woodchip treatments, compared with the standard feed treatment: pre-feeding (p < 0.001), feeding (p < 0.001) and post-feeding (p < 0.001. When we looked at the change in visible eye white within each treatment, during the concentrates treatment the eye white increased during the feeding segment compared with both the pre-feeding and post-feeding segments (p < 0.001). The visible eye white also increased significantly in the feeding segment of the woodchip treatment compared with during the post-feeding segment (p < 0.001), but not compared to the pre-feeding segment (p = 0.25). There is a need for more comparable research to be performed that explores both types of valence and arousal levels, before the effects can be fully understood. With this information and understanding, it would then be possible for visible eye whites to be used as a non-invasive measure of emotional state.