This document presents a method for classifying ECG signals using a fuzzy Petri net with inherent fuzziness. The fuzzy Petri net structure is organized into a neural network called a fuzzy Petri network. Features are extracted from ECG signals using wavelet transforms. A best basis technique is used to select optimal features that reduce the dimensionality of input vectors and complexity. The fuzzy Petri network parameters are learned using backpropagation. The system was tested on 8 classes of ECG beats and achieved accurate classification.