Spatial and social cognition are two aspects of fish behaviour that have been subject to an increasing amount of research in recent years, but few have investigated potential behaviour where the two coincide. Testing the ability for an individual to socially learn a spatial task would bridge this gap in understanding. We made naïve goldfish, Carassius auratus, observe a trained conspecific navigate a T-shaped maze, and then recorded how many trials it took for them to learn the maze, as well as time taken per trial, motivation and acceptance of the food reward. We also conducted reverse trials to understand whether allocentric or egocentric navigation was being learnt. On average, it took significantly longer for the observer group to learn the maze than it did the control group. The observer group were more likely to navigate allocentrically, whereas the control group were more likely to navigate egocentrically. There was no difference between allocentric control and allocentric observer subjects, but the egocentric control learned the maze in significantly fewer trials than egocentric observers. Although the observer group took significantly more time per trial and were less motivated, they were significantly more likely to accept the food reward. Although social learning was taking place, which was apparent from their increased acceptance of the food reward, something about the social information was inhibiting the learning process, as well as influencing navigation strategy choice. In conclusion, the act of a goldfish observing a trained conspecific resulted in a slower rate of training and an increased chance that place learning is occurring. This study assists in developing our understanding of spatial memory in teleost fish, which further justifies the use of zebrafish and goldfish as model species in spatial memory experiments.