Background and Objectives: Prostate cancer is one of the most commonly diagnosed cancers in the male population and the fifth leading cause of cancer death worldwide in men as of 2022. One of the potential biomarkers that can predict the progression of the disease is the transmembrane adhesion molecule CD44s. The aims of this study were to determine the expression of CD44s in prostate cancer in the central tumor mass and in the tumor periphery of the disease and to compare it with the clinicopathological parameters (PSA, Gleason score, surgical margins, and biochemical recurrence of the disease) in patients treated with radical prostatectomy. Materials and Methods: The research was randomized retrospectively during the period from 2001 to 2006. Tissue microarrays of 121 archival acinar prostate carcinoma samples were immunohistochemically evaluated for CD44s expression. The immunoexpression was determined semiquantitatively, taking into account the percentage (0 (0–5%), 1 (6–24%), 2 (26–75%), and 3 (76–100%) and intensity of the membranous staining of the tumor cells (0 absent; 1 weak at 400×; 2 intermediates at 100×; 3 strong at 40×) and calculated to obtain a final score (0–3 were regarded as negative; 4–6 were regarded as positive). Results: For statistical purposes, we divided the tumors into two categories: Gleason grade group 1 makes up 80.7% and grade group 2, which includes all the remaining Gleason grade groups (out of 2–5), accounts for 19.3% of the tumors. Grade group 1 had the highest incidence of score 4 (positive expression). There were statistically significantly more positive expressions in those tumors with negative prostatectomy margins (chi square: p = 0.001; Cramer V: 0.319). There was no correlation between CD44s expression and biochemical recurrence (p = 0.218), nor with the preoperative PSA values (p = 0.165). In the grade group 1 tumors, the CD44s immunoexpression and status of prostatectomy margin were statistically significantly related with negative margins (p = 0.028). An analysis of the expression of CD44s according to the localization in the central part of the tumor mass and on the periphery of the cancer in the group of tumors with a positive margin did not show a significant correlation because the sample was too small. Descriptively, it can be noted that the expression on the periphery was higher, and the central/peripheral expression ratio was higher in favor of the periphery. Conclusions: Our results provide insight into the possible value of CD44s expression for predicting the behavior of prostate tumors and the justification of therapy after a prostatectomy. Also hypothetically, they indicate a protective role of CD44s in a group of well-differentiated tumors at the periphery of the tumor mass. Therefore, it is useful to study the CD44s molecule further in this sense.