An assessment of the nutritional status of a representative sample of an urban African population has not previously been conducted, nor the extent to which the traditional diet has been abandoned for a western diet. To meet this end, a cross-sectional analytic study was carried out on a representative sample (N=1146) of the urban African population, aged 3 - 64 years in 1990. Particular attention was paid to specific at-risk groups viz. preschoolers (aged 3 - 6 years; N=163), adolescents (aged 15 - 18 years; N=119) and adults (19 - 44 years; N=649). The interrelationships of dietary intake with socio-economic status, demographic indicators as well as measures of urban exposure were also examined. A further aim was to determine the extent to which this rapidly urbanising population ' s macronutrient profile had shifted from a traditional towards a western atherogenic dietary pattern. This analytic study was nested in a community-based descriptive survey on risk factors for cardiovascular disease. A multi-staged, proportional sampling strategy was used. Quotas were used in the final stage of sampling, based on the age/sex distribution of a 1988 census conducted by the local authorities. Dietary data were collected by means of the 24-hour recall method, by Xhosa -speaking registered nurses who had received intensive training. Anthropometric measurements were taken, and blood samples were drawn according to standard procedures. Socio-demographic questions elicited information on the physical environment and facilities, educational level and employment status. Information was also elicited regarding urban exposure relating to lifetime migration history, thus incorporating retrospective temporality into the study des ign. From these data, an index of urban exposure was established by calculating the percentage of life spent in an urban environment. Univariate analyses of dietary, anthropometric and biochemical vitamin status were used for the descriptive components of the study of the three specific at-risk age categories. Bivariate analyses examined the effects of selected proxies of socio-economic status, and urban exposure on dietary intake. Finally, multiple linear regressions were performed on the preschoolers (N=163) and adult sample, aged 15 - 64 years (N=983) incorporating additional indicators of socio-economic status as predictors, and dietary intake data as outcome measures. Correspondence analysis further explored the relationships between dietary atherogenicity (using the Keys score) and other risk factors for degenerative disease.