Different machine learning methods have been developed so far in order to identify different stages of the disease Lupus. The disease Lupus is a complicated auto immune disease and treatment was developed during last few years. So, very less research work has been done in this field and few data sets are available to perform and test research work in this field. Due to the fact this domain is devoid of efficient methods that are capable of finding out various stages of lupus. This research work has made lupus data set from individual lupus patient using survey method. The proposed model has been used for identifying stage of lupus patient. A multi class perceptron has been developed to classify the stage of lupus patient. Six features have been considered as input neurons of proposed model. The stages of lupus have been considered as Normal, Moderate and Severe. Results have been compared to other neural network models. The performance and accuracy of the proposed model have been found better compared to other models. No existing model with limited data set could ever identify the stage with better accuracy. The proposed model is helpful for medical field as an alternative support system to identify the stage of lupus patient.