PURPOSE: To determine the usefulness of transrectal ultrasonography (TRUS) in diagnosing prostate cancer bycomparing the sensitivity, specificity, accuracy, and positive and negative predictive values of TRUS with thoseof serum prostate-specific antigen (PSA), prostate-specific antigen density (PSAD) and digital rectal examination(DRE). MATERIALS AND METHODS: Two hundred and ten consecutive patients underwent TRUS-guided prostate biopsy dueto elevated PSA and/or abnormal findings on TRUS or DRE. The TRUS findings were analyzed and correlated withpathological diagnosis. PSAD was calculated by dividing the serum PSA level by the prostate volume calculated onTRUS. The sensitivity, specificity, accuracy, and positive and negative predictive values of TRUS were comparedwith those of PSA, PSAD and DRE. Using ROC curve analysis, the combinations of these diagnostic methods were alsoevaluated for the determination of efficacy in diagnosing prostate cancer. RESULTS: The sensitivity andspecificity of serum PSA (cut-off level, 4ng/ml), PSAD (cut-off level, 0.15ng/ml/cm3), DRE, and trus were 96%/17%,96%/37%, 72%/62%, and 89%/68%, respectively. On TRUS, the sensitivity and specificity of low echoic lesions andthose of irregular outer margin were 89%/69%, and 60%/90%, respectively. TRUS was statistically more accurate thanother diagnostic methods. Of the combinations of diagnostic methods, TRUS and PSAD were most accurate. CONCLUSION: TRUS demonstrated lower sensitivity but higher specificity than PSA or PSAD. Although it is an accurate modalityfor the diagnosis of prostate cancer, it cannot be used as a confirmative test due to its relatively low positivepredictive value. A combination of diagnostic methods and random biopsy is needed in patients in whom prostatecancer is suspected.