This paper presents a multi-dimensional thermo-electrical model for pouch-type large-format lithium-ion batteries with high accuracy both spatially and temporally. The model includes a two-dimensional electrical model to calculate the current density over the current collector and a three-dimensional thermal model to calculate the temperature in the battery core. To improve the accuracy, key parameters are determined from carefully designed experiments, rather than taking from the literature. Thermal parameters are estimated using a joint experimental and computational method. Electrical contact resistance between the cable and the tab is estimated from a self-heating experiment. With the reliable input parameters, the predicted temperature evolution at multiple locations on the cell surface agrees well with the measured results, indicating high accuracy of the model. Using the validated model, sensitivity analysis is conducted to evaluate the relative importance of thermal parameters on battery thermal performance. It is found that the specific heat capacity has the greatest influence on the maximum temperature rise, while the in-plane thermal conductivity has the greatest influence on the maximum temperature variation. It is elucidated that the main cause for the temperature non-uniformity is the heat flux from the tab, rather than the non-uniformity of heat generation rate in the core.