The aim of this experimental research is to evaluate resistance training by up-to-the moment recording of mouse activities. Statistical population: Male mice (C57BL6-10 weeks old approximate weight 18-25 g) were randomly selected. The samples exercised in two groups: A (no light stimulus, 12 mice) B (with light stimulus, 15 mice) using a variable angle ladder with slope movement in 30°, 60° and 85°. The device consisted of four separate channels with a variable angle ladder at the bottom and movement sensors on each channel wall that recorded and drew the mathematical graph model of every training mouse activity separately using a particular hardware and software. Analyzing graphs showed that in the assumed exercise – training protocol members of one group in the same condition had not exercised similarly and their exercising model was significantly different by virtue of speed, time and intensity graphs. This difference was observed even in various training days of a mouse in comparison with itself. Therefore, the research indicated that it is impossible to assume the training activity of a group of mice in a resistance training to be the same even in a similar condition. As a result, the precise way of separating and grouping the trained mice is to monitor the training momently. With this method the effect of implemented training on each mouse is recorded and analyzed independently from others.