Airborne integrated path differential absorption (IPDA) lidar system is an important instrument to verify the performance and data inversion methods of future space-borne lidar systems for atmospheric CO2 measurement. A ground vertical path validation experiment of atmospheric CO2 measurement by an airborne double-pulsed 1.57-μm IPDA lidar has been implemented. The experiment was carried out and temperature, pressure and humidity profiles of Local Meteorological Station at almost the same time are adopted. Backscattering signals from clouds at altitudes of nearly 5 km were received. To avoid the influence of stray light from mirrors, the energy monitoring signal was delayed through the 200 m multimode fiber. But it is interfered by the aerosol scattering echo signals. Inversely, considering the stray light as monitoring signal, the inversion result of XCO2 is pretty good. Six methods are studied and compared to reduce the bias and improve the CO2 column-averaged dry-air mixing ratio (XCO2) accuracy. The “PIM, AVD” and “PIM, AVX” methods are more effective when clouds are acted as hard target. The mean value of lidar measured XCO2 calculated by “PIM, AVD” and “PIM, AVX” methods is 409.63 ppm. The average value of in-situ instrument UGGA is 411.05 ppm over the same period. The bias between IPDA lidar and UGGA is -1.42 ppm. With averaging 148 shots, the standard deviation of XCO2 of the IPDA lidar system is 3.68 ppm.