Rice (Oryza sativa L.) is one of the world’s most important cereals for human consumption. Different rice varieties vary significantly in their physical characteristics, chemical composition and mineral concentrations which may lead to different health implications for rice consumers. This is potentially important in Cambodia, where rice is the major source of carbohydrate intake in human diets. White rice intake has been associated with increased risk of developing type 2 diabetes, as reported by several large scale human observational studies. Starch, the main component in rice, has significant effects on the digestibility of cooked rice grains. The main objectives of this thesis are: i) to understand the rice consumption patterns of the Cambodian people across different agro-climatic zones, ii) to examine the relationship between starch structure and the digestibility of Cambodian cooked rice, iii) to investigate the effects of variety and growth location factors on grain composition and starch structure of Cambodian rice, and iv) to assess the effects of possible changes in rice consumption patterns on the burden of type 2 diabetes in Cambodia. A quantitative cross-sectional survey of a representative sample of 250 rural and urban households in five different agro-climatic zones of Cambodia was carried out to investigate rice consumption patterns. Characterization of molecular structure of debranched and whole starches in a selection of commonly consumed rice varieties was performed using size exclusion chromatography (SEC). The average molecular size and the degree of branching (DB) of whole starch molecules were analysed using offline multi-angle laser light scattering (MALLS) and 1H nuclear magnetic resonance (NMR) spectroscopy, respectively. An in-vitro method was used to study the starch digestibility of cooked rice grains. The effect of making selected changes to white rice consumption patterns on the incidence of type 2 diabetes in Cambodia was modelled using Prevent Plus software, using parameters based on the best available international evidence and integrating epidemiology and demographic principles. The results from the survey highlighted the diversity in varieties of rice grown and consumed in Cambodia. Thirty four and 36 rice varieties were reported to be grown and consumed by the respondents, respectively. Participants usually consumed rice between 2 and 3 times per day with an average amount of 462 g of uncooked rice per person. Rice consumption habits were found to be different across gender, location, season and types of consumer. The common important attributes for decisions on purchasing rice for consumption were price, taste and texture. The top ten rice varieties identified in the survey differed significantly in their starch structure and digestibility. A correlation analysis indicated a negative correlation between in vitro digestion rate coefficient, k, and amylose (Am) content. Positive correlations were found between k and DB and the average size of whole starch (Rg). Fine structural features of Am molecules influenced the digestion rate of starch in cooked rice grains. Those features which increased the digestion rate coefficient (more rapid digestion) were a smaller size of Am in fully branched starch molecules (hydrodynamic radius Rh at the Am peak maximum), a longer chain length of long Am branches (XAm2) and a smaller relative amount of short Am branches to short amylopectin (Ap) branches (hAm1/Ap1). Three among the ten popular rice varieties, collected from three different agro-climatic zones of Cambodia, were selected based on the difference in their Am contents to study the effects of variety and growth location on chemical composition and starch structure. Protein and lipid contents, but not the starch content, were affected by both variety and growth location. Crude protein content in rice grains increased when the rice was grown in a location with higher temperature, whilst crude lipid content increased when the rice was grown in a location with lower temperature. The fine structures of both Am and Ap were systematically different among varieties, but not affected by growth location in terms of the slight variations in temperature and rainfall. Systematic reviews of the literature have shown dietary glycaemic index (GI) and glycaemic load (GL) to be positively associated with the risk of developing type 2 diabetes. The GI of rice is known to be relatively high compared to other starchy foods. Data on rice consumption patterns and digestibility indices of two rice varieties, Phka Rumduol (high GI=88) and IR 66 (low GI=54), were used to model the potential effects of changing rice variety and amounts consumed on the burden of type 2 diabetes over a 10-year period. Projecting existing trends, the estimated diabetes burden with no intervention would increase from 11,315 new cases per year in 2011 to 14,852 in 2020. However, if the population changed from Phka Rumduol to IR66 this was predicted to be reduced by 27% and if the population consumed less rice (25% from the current consumption level) the increase in number of new cases was predicted to drop by 26%. The results of the thesis provide new evidence concerning the potential public health impact of rice starch structure-property relationships, which have important implications for food production and consumption policies in Cambodia as well as other rice-consuming countries. These findings should also encourage scientists to develop rice varieties with better nutritional characteristics to reduce the onset of type 2 diabetes and other related chronic diseases.