Changes in the immune system induced by tyrosine kinase inhibitors (TKI) have been shown to positively correlate with therapy responses in chronic myeloid leukemia (CML). However, only a few longitudinal studies exist and no randomized comparisons between two TKIs have been reported. Therefore, we prospectively analyzed the immune system of newly diagnosed CML patients treated with imatinib (n = 20) or bosutinib (n = 13), that participated in the randomized BFORE trial (NCT02130557). Comprehensive immunophenotyping, plasma protein profiling, and functional assays to determine activation levels of T and NK cells were performed at diagnosis, 3, and 12 months after therapy start. All results were correlated with clinical parameters such as Sokal risk and BCR-ABL load measured according to IS%. At diagnosis, low Sokal risk CML patients had a higher frequency of cytotoxic cells (CD8 + T and NK cells), increased cytotoxic potential of NK cells and lower frequency of naive and central memory CD4 + T cells. Further, soluble plasma protein profile divided patients into two distinct clusters with different disease burden at diagnosis. During treatment, BCR-ABL IS% correlated with immunological parameters such as plasma proteins, together with different memory subsets of CD4+ and CD8 + T cells. Interestingly, the proportion and cytotoxic potential of NK cells together with several soluble proteins increased during imatinib treatment. In contrast, no major immunological changes were observed during bosutinib treatment. In conclusion, imatinib and bosutinib were shown to have differential effects on the immune system in this randomized clinical trial. Increased number and function of NK cells were especially observed during imatinib therapy.