Simulation-based computer-assisted learning (CAL) is emerging as new technologies are finding a place in mainstream education. Dynamically linked multiple representations (DLMRs) is at the core of simulation-based CAL. DLMRs includes multiple visual representations, and it enables students to manipulate one representation and to immediately receive feedback from others. An interesting and important research question is whether verbalisers, who prefer to process verbal material, have similar learning performance and learning features as visualisers, who prefer to process visual material. To answer this question, 28 undergraduate students were selected as participants from the 855 undergraduate students who were initially tested with the style of processing scale (SOP). They were representative of either visualisers or verbalisers (students who scored upper 10% and lower 10% on the SOP). A study was conducted using an experimental design that included pre-and posttest and thinking-aloud methods. Simulation-Assisted Learning Statistics (SALS) was adopted as the learning environment for both groups. The analysis results are based on the data of 25 participants because three participants had trouble thinking aloud while using SALS. The results indicated that the visualisers and verbalisers did not differ significantly in their learning performance, but they did exhibit significantly different learning features in their use of DLMRs, their methods of reading learning guides and their learning strategies. Additionally, the learning features of the verbalisers explained why their learning performance was similar to that of the visualisers. Finally, this study provides recommendations for future applications and studies of simulation-based CAL.