There is little age-specific information on changes in dietary intake over time in this country, yet this is valuable in assessing the effectiveness of health education programmes particularly in vulnerable groups such as adolescents. In 1990, 379 children aged 12 years completed two 3 d dietary records. They were interviewed by one dietitian on the day after completion of each diary to verify and enlarge on the information provided and, with the aid of food models, obtain a quantitative record of food intake. Nutrient intake was calculated using computerized food tables. These children attended the same seven Middle schools in Northumberland as 405 children of the same age who recorded their diet using the same method, 10 years previously. Heights and weights were also recorded in both studies in the same manner. Comparing the nutrient intakes in 1990 with 1980, energy intake fell in the boys (to 8.6 MJ) but not in the girls (8.3 MJ). The contribution of fat to energy intake was unchanged at about 40% (about 90 g/d). Likewise, intake of sugars was unchanged at about 22% of energy (about 118 g/d). Calcium intake remained the same in the girls (763 mg/d in 1990) but fell in the boys (786 mg/d in 1990). Iron, vitamin C and unavailable carbohydrate intakes increased in both sexes, and the nutrient density of the diet improved in all sex and social-class groups. However, a social trend evident in 1980 still existed in 1990 with low social groups having the poorest-quality diet. It is concluded that there is little evidence of substantial progress towards improving the diet of adolescents in this country.