Resistance welding is a suitable technique for joining thermoplastic composites. Like other fusion bonding processes, it involves heating, melting and cooling steps. Productivity depends on the time that passes during these steps. This is the first study that tries to increase the productivity of the process in a systematic way. The objective of the present study is to determine the optimum set of process parameters to minimize the processing time. In order to ensure that the resulting joint satisfied the requirements of quality, the relationship between process variables and quality of the welded joint was established through process modeling. First, a one-dimensional transient heat transfer analysis was carried out using a finite difference method to find the temperature profiles across the thickness of the welding stack. Then, the heat transfer analysis was coupled with a degradation kinetics model in order to determine whether the resulting part has undergone excessive thermal degradation, or not. Finally, a bonding model adapted to the resistance welding process was used to determine the degree of bonding between the laminates. The process model was eventually combined with an optimization algorithm to minimize the processing time without violating the quality requirements. The algorithm was based on a search method called Nelder–Mead. Finally, optimum process parameters were obtained for different thicknesses of APC-2 laminates.