This study aimed to examine the association between meal-specific food patterns and incident hyperglycaemia in a Chinese adult population. Adults aged 20 years and older (n 1056) were followed from 2002 to 2007. Dietary data were collected using a 3-d food record and meal-specific (breakfast, lunch and dinner) food patterns were independently described by factor analysis based on the consumption of thirty-five food groups at each eating occasion. Each food pattern score was recoded as quartiles. Hyperglycaemia was defined as fasting plasma glucose >5·6 mmol/l at baseline and follow-up. The associated between food patterns and incident hyperglycaemia was assessed by logistic regression. During the follow-up, 125 new cases of hyperglycaemia were identified. Traditional (wheat) breakfast was inversely associated with incident hyperglycaemia, whereas traditional (rice, vegetable and pork) lunch and dinner were positively associated with the risk of incident hyperglycaemia, even after adjustment for a number of covariates including glycaemic load, carbohydrate intake and BMI. Incident hyperglycaemia occurred in 15·9, 13·6, 11·7, 6·1 % across quartiles of traditional breakfast; and 5·3, 9·1, 15·9, 17·1 % of the quartiles of traditional lunch pattern. The adjusted OR for hyperglycaemia was 0·67 (95 % CI 0·48, 0·92), 1·83 (95 % CI 1·32, 2·53) and 1·39 (95 % CI 1·04, 1·86) for 1 sd increase of traditional breakfast, lunch and dinner pattern factor score, respectively. A traditional wheat-based breakfast is associated with a decreased risk of hyperglycaemia. A rice-based traditional lunch and dinner is associated with an increased risk of hyperglycaemia in Chinese adults.