Antimicrobial de-escalation (ADE) is defined as the discontinuation of one or more antimicrobials in empirical therapy, or the replacement of a broad-spectrum antimicrobial with a narrower-spectrum antimicrobial. The aim of this review is to provide an overview of the available literature on the effectiveness and safety of ADE in critically ill patients, with a focus on special conditions such as anti-fungal therapy and high-risk categories. Although it is widely considered a safe strategy for antimicrobial stewardship (AMS), to date, there has been no assessment of the effect of de-escalation on the development of resistance. Conversely, some authors suggest that prolonged antibiotic treatment may be a side effect of de-escalation, especially in high-risk categories such as neutropenic critically ill patients and intra-abdominal infections (IAIs). Moreover, microbiological documentation is crucial for increasing ADE rates in critically ill patients with infections, and efforts should be focused on exploring new diagnostic tools to accelerate pathogen identification. For these reasons, ADE can be safely used in patients with infections, as confirmed by high-quality and reliable microbiological samplings, although further studies are warranted to clarify its applicability in selected populations.