We use the new external cavity widely-tuned quantum cascade laser (ECQCL) as the excitation light source and the high frequency quartz resonant tuning fork as the photoelectric detector to study the quantitative analysis and composition identification of volatile organic compounds (VOCs) by mid-infrared laser spectroscopy. The infrared absorption spectra of different types of gasoline samples from different sites are carefully analyzed and compared with the standard spectra taken from the Pacific Northwest National Laboratory (PNNL) database. The results are very consistent. In this experiment, a self-established spectral analysis model integrated with one dimensional cubic spline interpolation algorithm (CSIA) and multiple linear regression algorithm (MLRM) is developed and successfully used for quantitative analysis and component identification of VOCs.