The study was done in 1982/83 as an oral health baseline for a comprehensive health project involving primary school children in rural areas of Tanzania. A number of 399 standard two pupils with a mean age of 10.9 yr were dentally examined and interviewed using closed questions. Interexaminer reproducibility for caries was 99%, inter-interviewer reliabilities for population correspondence were between 81% and 100%. DMFS + defs was 1.1 for the 92% brushing their teeth daily and 1.9 for children brushing once a week or less. The amount of plaque, calculus, and gingivitis showed no relation to brushing, if it was performed more than once a week. The effect of using chewing sticks or plastic toothbrushes was identical. Children eating cakes, biscuits and/or sweets at least three times a day seemed to have more caries than the rest, DMFS + defs being 1.5 and 1.1, respectively. Consumption of sugar cane did not show any relation to the caries amount. In one of five districts traditional removal of primary tooth germs had caused a mean loss (or deformity) of 1.0 permanent tooth, thereby being the most injurious dental condition. The study confirms the need of educational interventions. It does not indicate any justification for prompting children in general to brush more than once a day or restricting sweet snacks to less than once a day.