<p>The existing researches on the flexibility evaluation and optimal scheduling of flexible loads in residential buildings do not fully consider the association characteristics of different loads, resulting in a large deviation between the calculated results and experimental results of optimization scheduling. A flexibility evaluation methodology and an optimization model considering load associations characteristics are proposed for flexible loads in residential buildings. Temporal flexibility ratio, which is the ratio of temporal flexibility considering association characteristics to that without considering association characteristics, is defined in this study. The optimization model is solved using the CPLEX solver under three different scenarios, namely, a scenario only considering the temporal overlapping load associations, a scenario only considering the temporal non-overlapping load associations, and a scenario considering both types of load associations. It was shown that in the residential building case in this study, the cooking loads with association characteristics exhibit less temporal flexibility but higher temporal flexibility ratio of up to 71.21%, while laundry loads exhibit higher temporal flexibility, but their temporal flexibility ratio is only around 36.84%. Additionally, when the users adopted the time of use (TOU) price, their electricity costs under the three considered scenarios increased by 0.00%, 7.57%, and 7.57% relative to the scenario without considering load associations, respectively. When installing a 3-kW household photovoltaic system, the electricity costs under the three scenarios increased by 0.00%, 1.28%, and 1.28%, respectively. As highlighted in the results, temporal non-overlapping association characteristics greatly affect the optimal scheduling of flexible energy loads, especially under TOU, while temporal overlapping association characteristics have little effect on that.</p>