An optimization method for configuring the regeneration, heating, and cooling sections of a plate and frame heat exchanger (PHE) in pasteurization processes is presented. The objective is to select the configuration with minimum operational and capital costs that satisfies constraints on fluid pressure drops, channel flow velocities, and pasteurizer thermal performance. The configuration of each section of the PHE is defined by its number of channels, pass arrangement, feed connection locations, and type of flow in the channels. Because the steady-state PHE simulation model for generalized configurations is in algorithmic form, it is not possible to represent the optimization problem in the form of a mixed-integer nonlinear programming model. Therefore, a branching method is proposed for its solution. Using a structured search procedure, the proposed method is able to successfully determine the optimal and near-optimal configurations with a very reduced number of exchanger evaluations. An example of optimization for a milk pasteurization process is presented, and it is verified that only 154 thermal simulations were required to obtain the optimal configuration from a feasible region containing 2.36 x 10(8) elements.