Abstract: Radar beam's lifting and widening in the distance and the vertical nonuniformity of the natural distribution of water content in the atmosphere are important factors of radar detecting rainfall errors. During 1980s—1990s, the importance of vertical profile of reflectivity (VPR) to radar detecting rainfall is emphasized and the study on the VPR technology is conducted. There are two major aspects on the researches of VPR, one is the determination method of VPR, the other is the technique to correct radar precipitation estimation. After comparing the major features of three existing determination methods, i.e., parameterization, averaging and identification, it is found that the mean vertical profile of reflectivity (MVPR) has the advantage of simplicity and utility. Four algorithms of MVPR are obtained and compared with emphasis on investigating the determination method for mean vertical profile (MVPR); and the different functions of determining parameters, especially for MVPR, are demonstrated. Results show that the R/G ratio can be enhanced from 0.84 (0.86) to 0.93 (0.97) and the AD of radar rainfall estimation can decrease by 4% after the MVPR correction in the two precipitation cases.Among the four algorithms of MVPR, M11 is the best one in which the ratio profile takes the near groundlevel as the datum plane so that the feasibility of radar precipitation correction can be assured, and it has better representation owing to consideration of the same weighing of each bin in domain D. The MVPR algorithms relate to 8 attribute parameters, of which some values exert significant effect on the quality of MVPR. For example, the initial distance D_s in domain D obviously affects the profile shape below 1 km, therefore 1~5 km should be adopted; the distance outside D₀ domain D plays its role from near ground-level to 6 km and the smaller D₀, the smaller the more smoothing due to the beam widening, so 40~50 km should be adopted; the smaller the value of the vertical piecewise-height Δh, the more detailed the profile shape, so 250 m should be adopted. The evidence also indicates that when the exponent b in the different Z-R relationships are kept constant, the slope of the profile below 1.5 km will increases and the bright band of the profile will be more distinguishable with the decrease of the coefficient A.By analyzing VPR derived from the volume data in terms of VCP21 model of CINRAD/SA in Wuhan, it is found that the bright band locates at the height of 5 km and the radar precipitation estimation is generally not affected by the bright band when using 0.5° reflectivity PPI (or hybrid scan) within radius of 200 km. When echoes of major precipitation system in domain D, Z (R) below the bright band decreases with the height (dZ/ dh < 0), and these may be the causes why radars underestimate summer rainfall (especially for long distance) in the middle reaches of the Yangtze and it could be done by means of VPR to correct the radar precipitation estimation.