Antimicrobial resistance poses a major threat to human health worldwide and the implementation of antimicrobial stewardship programs (ASPs), including antimicrobial de-escalation (ADE), is a multifaceted tool for minimizing unnecessary or inappropriate antibiotic exposure. This was a prospective observational study of 142 non-Intensive Care Unit (ICU) patients with microbiologically documented infection who were initially administered empirical antimicrobial therapy and admitted to the medical wards of 6 tertiary-care hospitals in Greece from January 2017 to December 2018. Patients were divided into two groups, the ADE and non-ADE group, based on whether ADE was applied or not, respectively. Exploratory end-points were ADE feasibility, safety and efficacy. ADE was applied in 76 patients at a median time of 4 days (IQR: 3, 5). An increased likelihood of ADE was observed in patients with urinary tract (OR: 10.04, 95% CI: 2.91, 34.57; p < 0.001), skin and soft tissue (OR: 16.28, 95% CI: 1.68, 158.08; p = 0.016) and bloodstream infections (OR: 2.52, 95% CI: 1, 6.36; p = 0.05). Factors significantly associated with higher rates of ADE were clarithromycin administration, diagnosis of urinary tract infection (UTI), isolation of E. coli, age and symptoms type on admission. Mortality was lower in the ADE group (18.4% vs. 30.3% p < 0.1) and ADE was not significantly associated with the probability of death (p = 0.432). ADE was associated with favorable clinical outcomes and can be performed even in settings with high prevalence of multi-drug resistant (MDR) pathogens without compromising safety.