There exist simple theories to predict the strain of sandwich structures under a bending load and specific support conditions. Although theoretical predictions are highly effective, but the role of testing in validation of these theories and their progress is undeniable. The main goal of the present paper is to prepare a test instruction in order to minimize the error of data acquisition and their analysis. The effects of specimen thickness, place of strain gauge installation and kind of strain gauge was investigated. Characterization of mechanical properties of test specimens are another focus of the present study. Classic theories are used to validate the results of the tests. Based on the large number of tests conducted on the samples, the results show that an important source of error in between experimental and theoretical results may be incorrect installation of the strain gauge. A method to find this out, is to obtain the results of loading and unloading. The method of using the gauge factor of the strain gauge on the package gives better results. Using a circuit with three wires omits the error of wire resistance. The results of the tests on composite specimens obtained by TML Company strain gauges showed small differences with the results obtained by BLH Company strain gauges. The error between experimental and theoretical results in sandwich specimens increases by increasing the core thickness. The measurements obtained by strain gauge in compression is not reliable. An important result obtained was that a half-bridge installation of the strain gauge decreases the error so much.