Objective To compare the relationship between the enzyme‐linked immunosorbent assay(ELISA) reagent and West‐ern blot(WB) confirmation reagent for analyzing the quality lever of human T‐cell lymphotropic virus(HTLV) detection reagent . Methods The WB confirmation reagent was used to detect anti‐HTLV antibody in 156 human serum samples of ELISA prelimina‐ry screening positive .The ELISA cut‐off value(optimal value) was selected by using the two‐graph receiver operating characteristics (TG‐ROC) analytical method .The two‐by‐two table analysis was constructed to analyze the consistency of results detected by the two methods ,moreover the McNemar test was used to evaluate the consistency of detection results .The quality level of HTLV de‐tection reagent was comprehensively evaluated .Results Among 156 serum samples of ELISA preliminary screening positive ,only 40 samples were positive by the WB confirmation ,and other 116 samples were negative .The sensitivity and specificity of ELISA de‐tection reagent obtained by TG‐ROC analysis were 97 .5% and 45 .7% respectively ,the TG‐ROC test also indicated that the detec‐tion results had significant difference between ELISA and WB(P<0 .05) .By adjusting the cut‐off value ,the sensitivity and specific‐ity of ELISA were increased to 88 .8% (parametric method) .In the comparison of the parametric method and the non‐parametric method ,the obtained areas under the curve(AUC) was 0 .923 5(parametric method) ,their results were basically consistent .Conclu‐sion Although above results indicate that the detection results of ELISA reagent are different from those of WB ,but adjusting the cut off value can increase its sensitivity and specificity ,thus increases the reliability of diagnosis result .