Includes bibliography. / This study provides a provisional normative range of performance for the South African semi- and unskilled black and coloured population groups on the Trail Making Test. One hundred and six black and coloured semi- and unskilled workers from the Groote Schuur Hospital in Cape Town, satisfying the criteria of 8 years or less of formal education, with no neurological impairment or psychopathology, and some degree of literacy, were randomly selected. Parts A and B of the Trail Making Test were then administered and scored, and the scores reported in tabular form, broken up in terms of race, age and education. A statistical analysis (multiple linear regression) was applied to the data and additional information on the effects of age, education and sex on test performance briefly discussed in terms of previous research findings. The results of this study add weight to Dugmore's (1987) assertion that existing norms currently in use for the Trail making Test are invalid for these population groups. Limitations of the study and suggestions for further research were also briefly discussed.