Abductive reasoning is the process of finding the best explanation for a set of observations. As the number of possible observations and corresponding explanations may be very high, it is commonly accepted that working memory capacity is closely related to successful abductive reasoning. However, the precise relationship between abductive reasoning and working memory capacity remains largely opaque. In a reanalysis of two experiments (N = 59), we first investigated whether reasoning performance is associated with differences in working memory capacity. Second, using eye tracking, we explored the relationship between the facets of working memory and the process of visuospatial reasoning. We used working memory tests of both components (verbal-numerical/spatial) as well as an intelligence measure. Results showed a clear relationship between reasoning accuracy and spatial components as well as intelligence. Process measures suggested that working memory seems to be a limiting factor to reasoning and that looking-back to previously relevant areas is compensating for poor mental models rather than being a sign of a particularly elaborate one. Following, high working memory ability might lead to the use of strategies to optimize the content and complexity of the mental representation on which abductive reasoning is based.