Handwritten character recognition is one of the important issues in pattern recognition applications. The applications of digit recognition include postal mail sorting, bank check processing, data entry form, pin code identification, identify the doctors' prescriptions etc. The conversion of an image based on the digit into letter codes for further use in a computer or text processing application is the first step in an off-line handwriting recognition system. This paper presents an approach of (MLP) Multilayer Perceptron neural network to recognize and predict handwritten digits from 0 to 9. MLP is a class of feed-forward artificial neural network. It consists of at least three layers of nodes having an input layer, a hidden layer, and an output layer. Each node is a neuron that uses a nonlinear activation function. The MNIST database is a large database of handwritten digits that is commonly used for training various image processing systems contain 60,000 training images and 10,000 testing images. The dataset was trained using gradient descent back-propagation algorithm and further tested using the feed-forward algorithm. The system performance is observed by varying the number of hidden layers. Various algorithms used for image processing have been discussed. The main objective of this paper is to ensure effective and reliable approaches for recognition of handwritten digits.