Purpose The concentration of PSA (Prostate Specific Antigen) after radical prostatectomy in prostate cancer patients is a predictor of biochemical recurrence, and the AUA (American Urological Association) is defined as biochemical recurrence when the concentration of PSA is measured at 0.2 ng/mL or more, and when the concentration is measured at 0.2 ng/mL or more at the retest. This standard is also applied our hospital. In this laboratory, the PSA reagent using IRMA (Immunoradiometric Assay) is used, and the sensitivity at a very low value was not as good as the reagent used in the department of laboratory medicine. This study aims to increase the reliability of the results by improving the precision and sensitivity of very low values. Materials and Methods As a reagent for the study, PSA reagent using IRMA was used. As a method to improve the precision and sensitivity of very low values, a variation method on the serum volume(25 uL, 50 uL, 100 uL, 200 uL) was studied, and variation usefulness evaluation was conducted. The evaluation items were compared the results of precision, analytical sensitivity, recovery rate, dilution test, high-dose hook effect test, parallel test and very low concentration values(n = 20). Results The validation results were displayed in the order of 25 uL, 50 uL, 100 uL, 200 uL. As the serum volume increased, it was confirmed that CV (Coefficient of Variation)(%) improved. Analytical sensitivity(ng/mL) was 0.038, 0.041, 0.017, 0.015 and recovery rate(%) was 101±3, 101±3, 99±2, 97±4. very low concentration values(ng/mL) between each volume(n=20) were 0.135±0.068, 0.076±0.050, 0.048±0.034, 0.046±0.034. and high dose hook effect appeared as the serum volume increased. Conclusion Through the variation usefulness evaluation, it was confirmed that as the serum volume increased, the precision and sensitivity improved at very low concentration values. However, it is necessary to pay special attention to the occurrence of high-dose hook effect as the serum volume increases. In the case of tests that requires very low concentration values, it is thought that the reliability of the result will be increased if the variation method is properly used after the variation usefulness evaluation.