Efficient detection of estrus is a permanent challenge for successful reproductive performance in dairy cattle. At present, in most parts of the world, estrus detection is done with observational basis, which has low productivity and causes great losses to this industry. Among the methods used to automatically detect estrus the electronic nose can be used to detect estrus from milk based on its odor due to advantages such as high speed, low operating cost and low manufacturing cost. In this study, an electronic nose system consisting of 7 metal oxide semiconductor (MOS) sensors was used to detect and predict the time of estrus. In this study, milk of 34 cows was sampled for 30 days in the Livestock of the University of Tehran in Karaj-Mohammadshahr and was tested on the same day, 26 of them showed signs of estrus and thus were examined for analysis. The system was analyzed along with pattern recognition methods such as Principal component analysis (PCA), Linear discriminant analysis (LDA), and Artificial neural network (ANN). At first, the feature vector was extracted from the response signal of the sensors and after sorting and classification, it was used as the input of pattern recognition methods. The results of pattern recognition methods were obtained as (94%) for principal component analysis, (86.33%) for Linear discriminant analysis, and (91.80%) for artificial neural network. The results showed well that the electronic nose system is a very efficient and fast way to classify and separate the different stages of the estrous cycle, particularly estrus stage.