There is currently no biomarker that reliably predicts treatment‐free remission (TFR) in chronic myeloid leukaemia (CML). We characterised effector and suppressor immune responses at the time of tyrosine kinase inhibitor (TKI) cessation in patients from the CML8 and CML10 clinical studies. Natural killer (NK) cells with increased expression of activating NK receptors were higher in patients who achieved TFR. There was no difference in the proportion of CD4⁺ or CD8⁺ T cells. Furthermore, we found that FoxP3⁺ regulatory T cells (T reg) and monocytic myeloid‐derived suppressor cells (Mo‐MDSCs) were concomitantly decreased in TFR patients, suggesting that the effector and suppressor arms of the immune system work in concert to mediate TFR. A discovery cohort (CML10) was used to generate a predictive model, using logistic regression. Patients classified into the high‐risk group were more likely to relapse when compared with the low‐risk group (HR 7·4, 95% CI 2·9‐19·1). The model was successfully validated on the independent CML8 cohort (HR 8·3, 95% CI 2·2–31·3). Effective prediction of TFR success may be obtained with an effector‐suppressor score, calculated using absolute NK cell, T reg, and Mo‐MDSC counts, at TKI cessation, reflecting the contribution of both immune suppressors and effectors in the immunobiology underlying successful TFR.