Within the scope of an information campaign of the Belgian Institute for Road Safety an attempt was made to classify 179 medicinal drugs from 9 therapeutic groups, listed in the Belgian "Commented Repertory of Drugs--1997", according to their effect on driving performance. The categorisation was based on literature data from about 500 references and used the system proposed by Wolschrijn et al : 7 classes ranging from no effect (I) over minor and moderate (II.1,II.2) to severe effects (III), completed with the respective* categories (I*,II*,III*) for presumed classes with insufficient scientific data. Forty-two drugs (24%) were considered having severe effects (III/III*). Only 28/179 molecules (16%) were classed in I/I*: no hypnotics-sedatives (33), anticonvulsants (10), antidepressants (25), neuroleptics (29), nor narcotic analgesics and antitussives (18) were listed in this no-effect category, while for 7/24 antihistamines (5/20 H1 and 2/4 H2), 12/20 beta blockers and 9/10 central stimulants the effect was considered negligible. Antidiabetics were not classified, as the danger lies in the risk of hypoglycemia due to inadequate use. The classification of the molecules proved to be problematic due to the lack of study data (42% of molecules in presumed categories) and the diversity in the study protocols. The effect on driving ability is dose-dependent and time-related, which makes the use of a single category inadequate; the effect further depends on co-ingestion of other medicines or alcohol, the development of tolerance and the condition of the subject. Physicians and pharmacists can use the proposed categorisation as a scientific base for guiding their patients, but should take into account the factors involved for each patient when estimating the driving ability.