Methods for predicting weld properties based on welding parameters are needed in friction stir welding (FSW). FSW is a joining process in which the resulting properties depend on the thermal cycle of the weld. Buckingham's Pi theorem and heat transfer analysis was used to identify dimensionless parameters relevant to the FSW process. Experimental data from Al 7075 and HSLA-65 on five different backing plate materials and a wide range of travel speeds and weld powers was used to create a dimensionless, empirical model relating critical weld parameters to the peak temperature rise and cooling rate of the weld. The models created have R-squared values greater than 0.99 for both dimensionless peak temperature rise and cooling rate correlations. The model can be used to identify weld parameters needed to produce a desired peak temperature rise or cooling rate. The model can also be used to explore the relative effects of welding parameters on the weld thermal response.