Traditionally, distributed BOTDR technique is applied in the field of the structural heath monitoring (SHM) on the large scale. However, if the sensing length is relatively small, the number of valid sampling points on the sensing fiber will be quite limited. Therefore, how to pick out valid sampling points from a large amount of sampling points becomes a key to apply distributed BOTDR technique into the field of small-scale monitoring successfully. This paper presents the concepts of valid sampling points (VSPs), invalid sampling points (ISPs) and useless sampling points (USPs), proposes SRSI Method to select VSPs, derives the error theory of ISPs and analyzes the experimental data by SRSI Method, Peak Method and the error theory of ISPs in the temperature and strain monitoring tests. In the above tests, different sensing lengths of the fiber or spatial resolutions were set for comparison with each other. The test results show that SRSI Method, which can improve the accuracy of the test results effectively, is the better selecting method of VSPs than Peak Method, and especially, the number of VSPs on the sensing fiber is less, the SRSI Method is better. Meanwhile, the error theory of ISPs gives the error bound and a good explanation of the large errors of Peak Method.© (2011) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.