Objective: In order to realize the rapid detection of benzo[a]pyrene content in barbecue products, a method of detection based on mid-infrared spectrum model was proposed. Methods: The original spectra of 120 barbecue samples were collected by mid-infrared spectroscopy. The original spectra were pretreated by standard normal variable transformation, multivariate scattering correction, first derivative and second derivative. The pretreatment effect of the spectra was evaluated by principal component analysis. The original spectral data were divided into training set and verification set at a ratio of 8∶2 to establish the evaluation model of barbecue products. Results: The R2c, R2p, RMSEC, and RMSEP of the model based on mid-infrared spectrum data were 0.995 1, 0.995 7 and 0.164 4 μg/kg, and 0.116 3 μg/kg, respectively. The residual error of training set samples was small after modeling. The maximum value was 0.406, and the total residual sum tended to zero. Conclusion: This model has strong prediction ability and high accuracy, and can better reflect the content of benzo[a]pyrene in actual barbecue. Thus, this can help to realize the non-destructive and rapid detection of barbecue products and has practical application value in the industry of barbecue products.