This study aimed to investigate whether the use of a judgement bias task with differential reinforcement was suitable to assess emotional states (positive or negative affect) in domestic hens. Six hens were trained to discriminate between a vertical and horizontal rectangle using an operant procedure with differential reinforcement to each stimulus in order to induce a positive or negative affect. Three ambiguous rectangles were introduced at intermediate angles and responses to these and the two learned rectangles were used to measure for judgement bias. Initial judgement bias probes measured response bias using three different foods as reinforcement; poultry pellets, wheat and puffed wheat in a one food magazine operant chamber. The chamber was then modified from a single food magazine to a double food magazine allowing six different food combinations with the three foods and differing reinforcement periods. Judgement bias probes measured response bias to the learned and ambiguous rectangles using combinations of each food type. Overall response latencies revealed a consistent pattern across all six hens with the fastest response times to the vertical (positive) stimulus and slowest response times to the horizontal (negative) stimulus. Results indicate that judgement bias was best measured and identified in the one food magazine chamber with four of the six hens consistently demonstrating positive affect and two hens demonstrating slight negative affect. It was found that results varied in the double magazine chamber revealing both response bias and food preference.