This paper presents alternative approach in heat transfer analysis of plate heat exchangers. In order to obtain heat transfer rate and effectiveness values of plate heat exchanger, neural network (NN) approach was used. Experimentally, system used in plate heat exchanger for heating and cooling applications was designed and constructed. Experimental data were used for training and testing network. The training and validation were performed with good accuracy. The correlation coefficient obtained when unknown data were applied to the networks was 0.9994 for heat transfer rate and 0.9976 for effectiveness, which is very satisfactory. Using the weights obtained from the trained network, a new formulation is presented for determination of heat transfer rate and effectiveness. This formulation can provide simplicity in thermal analysis of plate heat exchanger. The presented procedure can also help to heat exchanger designer and manufacturer.