The detection of the exhaust gas components of a fuel engine has important reference values for engine condition determination and environmental pollution monitoring. For this study, a weedkiller engine powered by No.95 gasoline was chosen as the experimental prototype, and the exhaust gas from the engine was blown directly into the signal acquisition focus of the Raman integrating sphere spectrometer. The high gas detection limit and the ability of the Raman integrating sphere spectrometer to probe all molecular gas qualitatively and quantitatively have been used to probe the molecular component of the gas in the tails. The gas component detected in the trailing gas is dominated by N2, O2, CO2, CO, and unburned gas. The relative Raman characteristic peak intensities of O2 (1 553 cm-1), CO2 (1 285 cm-1 and 1 388 cm-1), CO (2 144 cm-1), and unburned gasoline (2 894 cm-1) were obtained by normalizing the Raman spectral intensities using the Raman characteristic peak intensity of nitrogen vibrations as a standard. It can be found that the characteristic peak of CO does not appear in the spectrum of volatile matter of air and gasoline, and the content of O2 and CO2 in volatile matter of gasoline does not change significantly compared with that of air, while the relative intensity ratio of Raman characteristic peak of CO2 Fermi formant 1 388 and 1 285 cm-1 change. The working state of a lawnmower is divided into idler, first and second gear. When operating, the O2 content in the exhaust components is all below that in the air, allowing quantitative analysis of the amount of O2 consumed during engine operation. The O2 content in the exhaust increases relatively when the fuel engine is increased from idle to first and second gear. It is because as the engine gear increases, so does the air intake, and the proportion of oxygen involved in engine combustion is relatively reduced. At the same time, the amount of CO2 in the tailpipe gas increases dramatically compared to the amount in the air, suggesting that the working process of the fuel engine produces large amounts of CO2. The proportion of CO2 in the tailpipe gas gradually increases as the gear is raised and the engine power is increased. One of the main sources of CO2, the main cause of the greenhouse effect, is the use of fossil fuels. The data shows a positive correlation between the amount of CO in the tail gas and the amount of gasoline in the tail gas, suggesting that when combustion is insufficient, there is more gasoline left, and the amount of CO as a product of insufficient combustion also increases. With increasing engine gear, the absolute intensity of the characteristic peak of N2 decreases due to the decrease of the Stokes scattering strength of nitrogen with increasing engine exhaust temperature. In this paper, the Raman integrating sphere spectrometer is used to analyze the changes engine exhaust composition under different conditions. Moreover, the relationship between engine state and gas concentration is preliminarily established. We explore the application of Raman integrating sphere technology to fuel engine exhaust detection and verify its feasibility.