Abstract: The authors selected five sheets with different geological settings and metallogenic conditions and adopted such methods as greater symmetrical spacings of gridded data, reduction of the spacing between the measuring points and sample compositing to conduct a qualitative analogue and quantitative comparison. They also analysed the extent of control of the measuring grid, geochemical trend changes of the survey area, distribution of the anomalies and their area, quantity, scale and intensity, and variation of zoning within the anomalies to calculate the probability of anomaly non-missing. The results show that it is appropriate to reduce every other point on the basis of an original 500×100 m grid. If a 500×300 m grid is used, there will be agreater geochemical trend and variation of the anomalies. The general regularity is as follows: the total area of the anomalies decreases with reducing grid spacing while the number of the anomalies increases. The results also show that it is the most suitable to use a 500×200 grid for sampling, wnich can be converted into a 10 points/km~2 control. Sampling on a 0.25 km~2 grid can reduce about 4500 samples for each sheet and save more than 20000 Yuán