Background: Multimorbidity, polypharmacy, and inappropriate prescribing are significant challenges in the geriatric population. Tools such as the Beers List, FORTA, and STOPP/START criteria have been developed to identify potentially inappropriate prescribing (PIP). STOPP/START criteria detect both potentially inappropriate medications (PIMs) and potential prescribing omissions (PPOs). The latest, third version of STOPP/START criteria expands the tool, based on the growing literature. The study aimed to evaluate the prevalence of PIP and the number of PIP per person identified by STOPP/START version 3 and to compare it to the previous version. Methods: This retrospective, cross-sectional study enrolled one hundred geriatric patients with polypharmacy from two day-care centers for partially dependent people in Poland. Collected data included demographic and medical data. STOPP/START version 3 was used to identify potentially inappropriate prescribing, whereas the previous version served as a reference. Results: STOPP version 3 detected at least one PIM in 73% of the study group, a significantly higher result than that for version 2 (56%). STOPP version 3 identified more PIMs per person than the previous version. Similarly, START version 3 had a significantly higher prevalence of PPOs (74% vs. 57%) and a higher number of PPOs per person than the previous version. The newly formed STOPP criteria with high prevalence were those regarding NSAIDs, including aspirin in cardiovascular indications. Frequent PPOs regarding newly formed START criteria were the lack of osmotic laxatives for chronic constipation, the lack of mineralocorticoid receptor antagonists, and SGLT-2 inhibitors in heart failure. Conclusions: This study showed the high effectiveness of the STOPP/START version 3 criteria in identifying potentially inappropriate prescribing, with a higher detection rate than version 2.