Yogurt is a food produced by bacterial fermentation of milk. All kinds of nutrition components are changing dramatically in the process of fermentation. Therefore, it is important to establish a fast and efficient measurement technology of yogurt nutrition, which is also an important goal for food safety supervision in terms of monitoring the yogurt production process in real time. Fourier transform infrared spectroscopy (FTIR) has been widely used in the field of food safety, for it has high efficiency, high throughput, no chemical pollution, thus it can be used in the inspection of food adulteration. Our study has established a quantitative model to predict the nutrition components in yogurt, such as energy value, protein, fat, carbohydrates and sodium content. Based on the least squares (PLS) method, the model used CaF2 film FTIR technology. The results show that the new model can be used in quality control of yogurt production process: The R2 values of the model were 0.938 9, 0.926 6, 0.918 6, 0.941 8 and 0.977 1, comparing energetic value, protein, fat, carbohydrate and sodium contents with the original spectrum of calibration samples by cross validation. And the predictive R2 are 0.920 5, 0.905 3, 0.908 5, 0.939 3 and 0.936 4 respectively. Thus, the model has good prediction accuracy and reliability, which provides a feasible method for the rapid measurement of yogurt quality. As a preliminary exploration of the quality control technology of dairy products, this method has a good prospect of application.