Recently, absolute lymphocyte count (ALC) at diagnosis, as a surrogate marker of host immunity, has been reported to be a prognostic factor for clinical outcomes in extranodal NK/T cell lymphoma (ENKTL). In this retrospective study, we set out to investigate whether ALC at the time of confirmed relapse or at last follow-up is a marker for relapse after chemoradiotherapy in 84 patients with stage I/II ENKTL. Receiver operating characteristics (ROC) curve and area under the curve (AUC) analysis showed that ALC at follow-up was a significant marker for relapse (AUC = 0.883, P < 0.001). Using 1.215 x 10(9)/L as the optimal cutoff value of ALC, 44 patients (52.4 %) were in lower ALC group and 40 patients (47.6 %) were in higher ALC group. The sensitivity and specificity for ALC at the time of confirmed relapse or at last follow-up was 94.1 and 76.0 %, respectively. The relative risk of relapse with an ALC < 1.215 x 10(9)/L was 14.5. The positive predictive value with an ALC < 1.215 x 10(9)/L was 72.7 %, and the negative predictive value with an ALC a parts per thousand yenaEuro parts per thousand 1.215 x 10(9)/L was 95.0 %. The 4-year cumulative incidence rate for an ALC < 1.215 x 10(9)/L was 73.2 % compared with 3.2 % for an ALC a parts per thousand yenaEuro parts per thousand 1.215 x 10(9)/L (P < 0.001). In a multivariate regression analysis, ALC at the time of confirmed relapse or last follow-up remained to be a significant factor for relapse (P < 0.001). In conclusion, lymphopenia observed during routine follow-up can predict relapse in patients with ENKTL, which needs further validation in prospective trials.