Extinction of learning is a common, yet under-reported limitation of judgment bias testing methods Repeated exposure to the ambiguous probe of a judgment bias paradigm encourages the animal to cease display of the required behaviours. However, there remains a need to repeatedly test animals to achieve statistical power. A delicate balance therefore needs to be struck between over- and under-exposure of the animals to the test conditions. This study presents the data of rats, a common animal subject of judgment bias testing. Rats were exposed to the ambiguous probe of a common, active-choice judgment bias test for 11 consecutive days. There was a significant increase in the latency to respond to the ambiguous probe following day 8, with no significant increase experienced for either the positive or less-positive probes. Following day 8 there was a significant increase in both optimistic and pessimistic latencies in response to the ambiguous probe. Therefore, repeated exposure to the ambiguous probe caused an increased latency in response even though optimistic interpretations were recorded. This implies that the use of response latency alone as a measure in judgment bias testing can falsely identify pessimism. Researchers should modify experimental design to include both choice and latency measures.