Aim: To assess the effect of physical training on the selected parameters of the immune system regarding CD3, CD4, CD8, CD11, CD161, CD45A cell counts in rats treated with N-methyl-N-nitrosourea (MNU). Material and Methods: Thirty-eight female Sprague-Dawley rats were injected intraperitoneally with MNU and were divided into three groups, i.e., sedentary control (SC), the group of moderate-intensity training (MIT) and the group of high-intensity training (HIT). Physical training was supervised immediately after MNU administration and was conducted 5 days per week for 12 weeks on a three-position treadmill. Results: A significant difference was found between SC and training groups in terms of the number of induced tumors per rat (1.57 vs. 0.4, p = 0.05) and in the following lymphocyte subpopulations: CD4+/CD8+ (p = 0.01), CD3−/CD11b+ (p = 0.02), CD3−/CD161+ (p = 0.002), CD3−/CD161− (p = 0.002), CD3+/CD45RA+ (p = 0.003) and CD3−/CD45RA+ (p = 0.005). In terms of the intensity of physical training, the highest efficacy was found for MIT and the following lymphocyte subpopulations: CD3−/CD11b+ (SC vs. MIT, p < 0.001), CD3−/CD161+ (SC vs. MIT, p = 0.002), CD3−/CD161− (SC vs. MIT, p = 0.002), CD3+/CD45RA+ (SC vs. MIT, p = 0.02) and CD3−/CD45RA+ (SC vs. MIT, p < 0.001, MIT vs. HIT, p = 0.02). Furthermore, negative correlations were found between the number of apoptotic cells and CD3−/CD11b (r = −0.76, p = 0.01) in SC and between the number of induced tumors and CD3+/CD8+ (r = −0.61, p = 0.02) and between their volume and CD+/CD8+ (r = −0.56, p = 0.03) in the group of rats undergoing training. Conclusions: Physical training, particularly MIT, affected immune cell function and an altered immune response can be considered a mechanism underlying the effect of exercise on breast cancer development.