Abstract: The gray error theory and its application in data processing of rock and mineral analysis are introduced in this paper. The results show that the accumulation values of the measurement data have linear relationship with the measurement times in rock and mineral analysis in which the measurement data are in a narrow error range. And the accumulation values of the measurement data have nonlinear relationship with the measurement times in the rock and mineral analysis with uneven element distribution (such as gold analysis) and with big dispersive data (such as rock intensity tests). But the standard deviation of the gray error theory is very close to the standard deviation of the traditional statistics. The method is simple, precise and useful. It is especially practical to the measurements with few data or difficult to follow the statistic regulation.