The spatial variation of crop yield was mainly caused by the complicated factors, such as soil and land factors. Based on sound theoretical reasoning, the paper performed an exploratory investigation of the relationship between the spatial distribution of yield measurement about 849 samplings on Thiessen polygon. Reasonable sample size or sampling density is an assurance of yield estimation surface. To investigate the spatial structure of Ji wheat 22 yield, we developed the comparison analysis of surface interpolation precision by stochastic samplings of 568, 425, 212 and 144 from a total of 849 samplings in zoning area, the other emphasis of the study was explored the influence of the distribution change of yield with the threshold of samplings numbers for the higher estimation precision. We first detected the presence of global spatial autocorrelation characteristics about the distribution of the Thiessen polygon region generated yield measurement samplings, then with the support of IDW interpolation, the higher precision surface was analyzed and obtained by the comparison of 5 samplings mode. High yield Thiessen polygon region of Ji wheat 22 tended to be clustered close to the high yield polygon region, the low yield Thiessen polygon region tended to be clustered close to the low yield polygon region. It is important to estimate the yield distribution for sampling density, samplings method, interpolation methods and sample size, we suggested that a suitable sample size of about 200 samplings should be used for Ji wheat 22 yield developed from the sampling zoning to acquire the higher surface estimation precision, beyond or the samplings number, the estimation precision presented the fluctuation characteristics.