A major obstacle in the pre-clinical study of mood-related disorders and novel affective state therapeutic evaluation is the lack of animal models that fully recapitulate human symptomatology. In this study, we developed a touchscreen-based Go/No Go cognitive judgement bias (CJB) task for mice. In this task, animals first learned to discriminate between two visual stimuli displayed on the touchscreen: one associated with a reward (S+) and one associated with a time-out and flashing house light (S-). Once mice learned to respond to the S+ and to withhold responding to the S- consistently, a set of four ambiguous stimuli ranging in visual similarity to the S+ and S- stimuli were randomly interspersed in the stimulus presentation sequence. Responses to these ambiguous stimuli were interpreted as a greater expectation of positive ('optimistic bias') or negative ('pessimistic bias') outcomes as a function of their similarity to the S+ or S- stimuli. The acute administration of the SSRIs fluoxetine and citalopram and the 5HT-2C receptor antagonist SB204048 did not produce any effects on CJB task performance. However, the noradrenaline/dopamine reuptake inhibitor, bupropion, increased responses to the ambiguous stimuli consistent with induction of an 'optimistic bias' and the pro-depressant tetrabenazine yielded the opposite effect. This study underscores the capacity of mice to respond to visually ambiguous stimuli in an ambiguity-dependent manner, a phenomenon observed across various species, including humans. Furthermore, it establishes and validates an operant behavioural task to assess CJB in mice delivered using the touchscreen platform which has significant intrinsic cross-species translational potential.