Medical students generally learn to calculate drug doses using paper problems containing numerical information. In contrast, once qualified, they have to extract the relevant information from treatment guidelines and use equipment, such as syringes, to dose patients. We compared students' success in calculating drug doses using those different approaches. After receiving relevant training, 239 consenting first year students were randomised into four groups. Groups 1 and 2 had to calculate a dose, while groups 3 and 4 had to both calculate and then prepare a dose using a syringe and a labelled bottle of medicine. We gave groups 1 and 3 numerical dosage information. Groups 2 and 4 had to extract this information from guidelines. We compared success between groups using a chi-square test. Groups 1 and 3 outperformed groups 2 and 4. Availability of equipment seemed to be related to the improved success between groups 2 and 4, but this was not significant. Students who did not speak English at home were three times more likely to answer incorrectly than English-speaking students. Students who had equipment were more likely than students who did not have equipment to use appropriate units and propose realistic quantities. In light of these findings, we recommend that medical student training and assessment should include both extraction of embedded dosage information from guidelines and use of the equipment used in dosing.