Neostigmine has been found to improve survival in animal models of sepsis. However, its feasibility, efficacy, and safety in patients with sepsis or septic shock have not been investigated. This parallel randomized controlled double-blinded design aimed to investigate the efficacy and safety of neostigmine as an adjunctive therapy in patients with sepsis or septic shock. A total of 167 adult patients with sepsis or septic shock were assessed for eligibility; 50 patients were randomized to receive a continuous infusion of neostigmine (0.2 mg/h for 120 h; neostigmine arm) or 0.9% saline (control arm) in addition to standard therapy. The primary outcome was the change in Sequential Organ Failure Assessment (SOFA) scores 120 h after therapy initiation. Secondary outcomes included mortality rates and changes in procalcitonin level. The median (interquartile range) change in SOFA scores improved significantly in the neostigmine arm -2 (-5, 1) as compared with the control arm 1.5 (0, 2.8); = 0.007. Progression from sepsis to septic shock was more frequent in the control arm ( = 0.01). The incidence of shock reversal in patients with septic shock was significantly lower in the control arm than in the neostigmine arm ( = 0.04). Differences in 28-days mortality rates did not reach statistical significance between the control and neostigmine arms ( = 0.36). Percentage change in procalcitonin levels was similar in both arms ( = 0.74). Neostigmine adjunctive therapy may be safe and effective when administered in patients with sepsis or septic shock. Clinical Trial Registration: NCT04130230.