This paper explores the use of mental maps as a diagnostic tool to assess university students' geographical knowledge of the Balearic Islands, by examining aspects that have been largely over-looked in this field of research, including certain formal characteristics of sketch maps. More specifically, it analyses whether required components (each island) and their correct location within the background context (the Balearic archipelago) are present, in addition to salient landmarks on the islands' coastlines. The quantitative procedures (a weights matrix of relative locations and a principal component analysis supporting cluster analysis) enabled the analysed maps to be classified and correlated. Failure to include one of the islands, inexact location and incorrectly traced outlines were the main indicators of students' lack of familiarity with the archipelago. In turn, the work highlights the correlation between formal aspects of mental maps and the level of geography knowledge. This finding underlines how useful sketch maps are as a tool to study geography knowledge, which is at odds with theories that question them without supporting empirical verification. Although drawing skills might slightly influence the results, this has no bearing on missing islands, headlands, bays and other features attributable to students' knowledge.