BackgroundThis study examines serum calcium levels and in-hospital mortality in patients with sepsis, a subject with contradictory findings in the existing literature. MethodsThis retrospective cohort study utilized data from the MIMIC-IV database, focusing on adult patients diagnosed with sepsis between 2008 and 2019. The serum calcium levels were taken as the highest value within the first 24 h of Intensive Care Unit (ICU) admission. We performed Cox proportional hazards regression analyses in multivariable-adjusted models to investigate the association between serum calcium levels and in-hospital mortality. Restricted cubic spline functions were used to assess the nonlinear relationship, and threshold effect analysis was conducted to identify potential inflection points. ResultsA total of 18,546 patients with sepsis were included in the study, and an in-hospital mortality rate of 16.9 % (3,126 out of 18,546) was obtained. Furthermore, a U-shaped relationship was observed between serum calcium concentrations and in-hospital mortality, with the lowest point at approximately 8.23 mg/dL. Hazard ratios were calculated as 0.75 (95 % CI: 0.67–0.85, P < 0.001) on the left side and 1.10 (95 % CI: 1.04–1.16, P < 0.001) on the right side of the inflection point. Sensitivity analyses corroborated these results. ConclusionThe research identified a U-shaped correlation between serum calcium concentrations and in-hospital mortality rates among patients with sepsis, underscoring the importance of serum calcium monitoring in this patient population upon hospital admission.