Objective: To determine the dietary intake, practices, knowledge and barriers to dietary compliance of black South African type 2 diabetic patients attending primary health-care services in urban and rural areas. Design: A cross-sectional survey. Dietary intake was assessed by three 24-hour recalls, and knowledge and practices by means of a structured questionnaire (n ¼ 133 men, 155 women). In-depth interviews were then conducted with 25 of the patients to explore their underlying beliefs and feelings with respect to their disease. Trained interviewers measured weight, height and blood pressure. A fasting venous blood sample was collected from each participant in order to evaluate glycaemic control. Setting: An urban area (Sheshego) and rural areas near Pietersburg in the Northern Province of South Africa. Subjects: The sample comprised 59 men and 75 women from urban areas and 74 men and 80 women from rural areas. All were over 40 years of age, diagnosed with type 2 diabetes for at least one year, and attended primary health-care services in the study area over a 3-month period in 1998. Results: Reported dietary results indicate that mean energy intakes were low (,70% of Recommended Dietary Allowance), 8086 -8450 kJ day 21 and 6967-7382 kJ day 21 in men and women, respectively. Urban subjects had higher ðP , 0:05Þ intakes of animal protein and lower ratios of polyunsaturated fat to saturated fat than rural subjects. The energy distribution of macronutrients was in line with the recommendations for a prudent diet, with fat intake less than 30%, saturated fat less than 10% and carbohydrate intake greater than 55% of total energy intake. In most respects, nutrient intakes resembled a traditional African diet, although fibre intake was low in terms of the recommended 3-6 g/1000 kJ. More than 90% of patients ate three meals a day, yet only 32 -47% had a morning snack and 19-27% had a late evening snack. The majority of patients indicated that they followed a special diet, which had been given to them by a doctor or a nurse. Only 3.4-6.1% were treated by diet alone. Poor glycaemic control was found in both urban and rural participants, with more than half of subjects having fasting plasma glucose above 8 mmol l 21 and more than 35% having plasma glycosylated haemoglobin level above 8.6%. High triglyceride levels were found in 24 to 25% of men and in 17 to 18% of women. Obesity (body mass index $30 kg m
22) was prevalent in 15 to 16% of men compared with 35 to 47% of women; elevated blood pressure ($160/95 mmHg) was least prevalent in rural women (25.9%) and most prevalent in urban men (42.4%). Conclusions: The majority of black, type 2 diabetic patients studied showed poor glycaemic control. Additionally, many had dyslipidaemia, were obese and/or had an elevated blood pressure. Quantitative and qualitative findings indicated that these patients frequently received incorrect and inappropriate dietary advice from health educators.