Abstract: Portable X-ray Fluorescence Spectrometer (PXRF) is often used in the laboratory and the field because it is a portable, high-efficiency and non-destructive piece of equipment. PXRF can be used to obtain semi-quantitative/quantitative results of multiple elements in geological samples within two minutes. However, there are many factors that can affect the analytical results of geological samples by PXRF, including the surface state, sample heterogeneity, and the measurement time. In order to further understand the effect of sample types and analytical methods on the element contents, the study described here compares the analytical results of rocks by PXRF with those acquired by conventional laboratory analysis. Results of rocks and rock powder samples by PXRF were also compared, and the effect of different detection time on the results of rock powder samples by PXRF was also investigated. A method has been proposed which will reduce the exploration cost and improve work efficiency without a significant loss of analytical precision. The results shows that analytical results of most of the detectable elements is not reliable when analyzing rocks directly by PXRF, especially for the frequently used trace elements (Cu, Pb, Zn, As and Ni). However, PXRF can be used to analyze rock powders and the analytical result is acceptable. There is a difference between the correlation of various elements between rock and rock powder samples by PXRF. Therefore, it is possible to determine whether the rock sample needs to be ground for sample preparation according to the target elements of interest in practical work. The detection time has no obvious effect on the elemental content results. Whether a specific element can be obtained in a relatively short time, further detection time is unnecessary.