The aim of the study was to evaluate the usefulness of blood vessel count as a diagnostic parameter in early detection of prostate cancer and to compare the number of blood vessels in pathohistological samples among the most common prostate diseases. This retrospective study included clinical and pathohistological data and samples from the Urology Clinic and the Department of Pathology in University hospital, Rijeka. Prostate cancer biopsies (n = 20), benign prostatic hyperplasia (n = 21) and prostatitis and nonspecific findings (n = 28) were treated immunohistochemically with CD31 antibodies. The subjects were also screened preoperative PSA values. The number of blood vessels were determined by counting the CD31 positive structures in the three areas of the largest number of blood vessels in each sample, we called the "hot spots". For the statistical processing of the results, the software package MedCalc Statistical Software version 14.8.1 was used The correlation between PSA values and the number of blood vessels has not been statistically significant. The univariate model in which the predictor of cancer was PSA was statistically significant (p = 0,0009) and was accurately classified 71, 67% of patients with prostate cancer. A multivariate model that included the number of blood vessels and the PSA values showed statistical significance (0.0001), and that model within value of p = 0.05 correctly classifies 80% of patients with prostate cancer. The number of blood vessels, as a diagnostic parameter, is not a good predictor of the diagnosis of prostate cancer, but due to its relatively high specificity, it represent a good complementary data to PSA value, whereby the number of correctly classified patients increases by 10%