The application of image recognition is widely found in the automation industry. Fundamentally, image recognition is a study on how the computer can recognize images automatically. Particularly, this research discusses about face recognition which is based on Wavelet Transform and Back-propagation Neural Network. A sample of 25 face images was taken from five people. All images were stored in the form of the 24-bit Bitmap format. The recognition process started with changing the images into 8-bit grayscale format, followed by extracting that image features with wavelet transform. By using this transform the size of the image was reduced from (96  128) pixels into (12  16) pixels. The result of this process was a matrix of [25  197]. Finally, the matrix was used as an input in the back-propagation neural network process for classifying the image. A number of experiments were then conducted to evaluate the performance of the system. The result has shown that this method has successfully recognized the given images with about 97.33% of accuracy and got threshold about 0.813108 which came from one hidden layer artificial neural network configuration.