In general, most original values of mechanical properties and life of products are obtained from reliability tests, but the reliability test is time-consuming and costly, and then appropriate sample size needs to be determined firstly when setting up a reli- ability test program. A probabilistic analysis method is presented to determine the minimum sample size with assigned precision in- dex, confidence level. The method is used for the cases where the test data follow Weibull distribution, and its shape parameter can be taken as known. Fortunately, the shape parameter of Weibull distribution for fatigue life of many materials used in aircraft struc- tures has been obtained from large amounts of fatigue-tested data through statistical approaches. A formula indicating the relation among sample size, precision index, confidence level and distribution parameters based on Weibull distribution is derived. As an example, the minimum sample size required to estimate the characteristic life (scale parameter) and the safe life (percentile of life) are calculated.