Abstract: Meteorological rocket sounding data are of great significance for research on the near-space environment. Based on two sets of meteorological rocket sounding data and L-band radiosonde data obtained in Bazhou, Xinjiang, in November 2023, this study evaluates measurement uncertainty using error analysis theory in conjunction with a temperature correction model, wind inversion model, and formulas for pressure and density calculation. A comparative analysis of the two types of sounding data was conducted, possible causes of discrepancies were discussed, and further comparisons were made with mainstream atmospheric models and numerical products. The results show that the uncertainties of temperature, wind speed, and wind direction measured by meteorological rockets are 0.3 to 2.8℃, 0.1 to 1.6 m/s, and 0.1° to 1.9°, respectively, while the relative errors of pressure and density are 1.1% to 4.1% and 1.1% to 4.3%, both increasing with altitude. Compared with L-band radiosonde data, the meteorological rocket data exhibit good consistency in variation trends, with smaller profile deviations and dispersion, capturing more measurement details with higher quality and accuracy. The main sources of deviation between meteorological rocket and L-band radiosonde data are differences in detection principles, temporal and spatial sampling, measurement errors inherent to the instruments, and model correction errors. The profiles from near-space environment forecasts and ERA5 reanalysis data align more closely with meteorological rocket data, showing smaller deviations. These results can support the subsequent calibration and application of models and data.