The smoothness of the end surfaces of a concrete specimen is an important factor influencing the result of the compressive strength test, and there is a possibility that the apparent compressive strength changes greatly due to the smoothness of the end surfaces. Therefore, it is important to grasp stress distribution on the loading surfaces during compressive tests.<br> Measurement of pressure distribution by pressure measuring films is performed by the color forming layer and the color developing layer in between the pressed part of the specimen and a loading plate. Color development is carried out by microcapsules of color forming layer when they are broken by the applied pressure, resulting in a chemical reaction in which a color former therein is adsorbed to the color developing layer. The film is considered guite useful to investigate and analyze the pressure and internal strains distribution in a hardened concrete.<br> In this study, pressure distribution on the upper end surface at various stages of loading of normal strength concrete and longitudinal strain of the middle height of the specimen were measured by using the pressure measurement films and strain gauges, respectively. Experiments were carried out in two series. In Exp. 1, various end face treatment methods were applied to the upper part of a specimen, and in Exp. 2, a step was formed on the upper end face of the specimen. We examined the influence of these experimental factors on the compressive strength and strain.<br> In addition, the reproducibility of the strain test results of the specimen by FEM (Finite Element Method) analysis using measured pressure distribution data was also examined.<br> As for various end surface treatments, difference in pressure distribution was clearly observed according to the difference in the end surface treatment method, but it hardly affected the averaged strains of opposite side of a specimen.<br> As for the specimens with the step gap on the upper end surface, it was found that the compressive strength was not greatly affected if the gap of step was within 0.1 mm, but was remarkably decreased when the gap exceeded 0.2 mm. There was no correlation between pressure distribution and strain distribution at the beginning of loading (stress stage of 5N/mm²). It is considered that the loading plate became oblique due to the step gap on the upper end surface, and the pressure applied to the specimen propagated diagonally.<br> The main results in this paper are as follows:<br> 1) Although there are differences in loading pressure distribution due to differences in end face treatment methods, the average values and of the opposite side strains Young's moduli are guile similar regardless the treatment methods.<br> 2) By using the average value of the two opposite side strains, the influence of the step on the upper end face on the Young's modulus can be greatly reduced.<br> 3) In the FEM analysis using the measured pressure distribution, the influence of the pressure distribution on the upper end face of the specimen was largely reproduced, in the strain distribution.