Emotions guide action in the light of evolutionary imperatives, and provide subjective meaning to experiences and events. The extent to which the range of emotional responses familiar to humans is shared with other species is unclear. This thesis aimed to further the comparative understanding of emotional states, and seek new tools for research into subjective emotion. This was achieved using experiments aimed at exploring induced emotional states in animals, including humans. Firstly, a novel test of anxiety-like affect was developed for three-spined sticklebacks. Based on scototaxis (dark preference) and novel tank diving, the successful use of this test indicates potential future utility across a range of fish species. It was used, along with open-field and novel-object tests, to assess sticklebacks' emotional responses to handling stress. These tests showed reduced preference for dark and deep areas of the tank, and reduced distance from the novel-object, following handling with a net rather than a scoop. Results indicate for the first time that acute stress can have an anxiolytic effect on fish. Handling stress was further used as an affect manipulation in development of a novel cognitive-bias test for fish. Human experiments explored potential mechanisms for manipulating cognitive and subjective components of mood independently. Evidence for an impact of viewing triangles of differing orientation was found on explicitly stated, but not implicitly measured, emotion. A test of facial interpretive bias was used along with subjective report to examine the effect of unpredictable (compared to predictable) sound presentation on anxiety. This found inconsistent effects on both cognitive bias and felt emotion, indicating that they are similarly sensitive to low level affect. The effects that were found included sex differences with greater responses in female participants. Results are discussed in relation to future work which could be carried out to distinguish conscious and nonconscious emotion.