Cognitive bias in decision-making under ambiguity is observed in animals and humans, and some individuals interpret ambiguous stimuli as positive (“optimistic”) while others as negative (“pessimists”). Social animals use both asocial and social cues to predict the presence of a reward or punishment in the environment through associative learning. In recent years, experimental paradigms have been developed to assess cognitive bias in various species based on their interpretation of ambiguous stimuli. The aim of this study is to determine whether a period of social isolation of adult zebrafish (Danio rerio) can affect judgement of an ambiguous stimulus, i.e. if it can bias the proportion of “optimistic”/”pessimistic” fish. We thus implemented a Go/No-Go task in a color-coordinated half-radial maze, where fish are trained to discriminate between a positive (P) and negative (N) arm. Discrimination learning is measured by the latencies to enter these arms. A Judgment Bias Score (JBS) is calculated based on latencies to enter the P, N, and an Ambiguous arm, allowing classification on an optimistic/pessimistic axis. In our preliminary results we observed two distinct behavioral phenotypes: active fish that explore all the maze and more static animals that either freeze or perform brief excursions. We also analyze how they interpret ambiguous stimuli and whether social isolation affects this cognitive process.