•Rice texture was evaluated by sensory panellists and two instrumental methods.•Two instrumental methods showed strong correlation with sensory results.•Dynamic rheological testing can be used to replace the conventional TPA method.•Rheological behaviour is explained from starch structural differences. Increasing demands for better instrumental methods to evaluate cooked rice texture is driving innovations in rice texture research. This study characterized cooked rice texture by descriptive sensory analysis and two instrumental methods (texture profile analysis (TPA) and dynamic rheological testing) using a set of 18 varieties of rice with a wide range in amylose content (0–30%). The panellists’ results indicated that hardness and stickiness were the two most discriminating attributes among 13 tested textural attributes. The consistency coefficient (K*) and loss tangent (tan δ) from a dynamic frequency sweep were used to compare with hardness and stickiness tested by TPA and sensory panellists, showing that using K* to express hardness, and tan δ to express stickiness, are both statistically and mechanistically meaningful. The instrumental method is rationalized in terms of starch structural differences between rices: a higher proportion of both amylose and long amylopectin branches with DP 70–100 causes a more elastic and less viscous texture, which is readily understood in terms of polymer dynamics in solution.