Background <p>Identifying high-risk acute pancreatitis (AP) patients in the ICU is vital for improving prognosis. Thus, this study aims to explore the relationship between the coefficient of variation (CV) of blood glucose and the all-cause mortality of patients with AP in the ICU.</p> Methods <p>A retrospective analysis was conducted on AP patients in the MIMIC-IV database. The CV was used to describe the glycemic variability (GV) and the optimal cut-off value was determined using the ROC curve. Subsequently, analyze the correlation between CV and all-cause mortality.</p> Results <p>A total of 907 patients with AP in the ICU were included in this study. The ROC curve determined the optimal CV cut-off value as 0.25. The KM survival curves and univariate and multivariate logistics regression analyses all showed that CV was associated with the 30-day, 60-day, and 90-day all-cause mortality (<i>P</i> &lt; 0.05). The RCS curves showed a nonlinear correlation (<i>P</i> &lt; 0.05). When CV is less than 0.421, 0.449, and 0.428, respectively, the risk of death at 30-day, 60-day, and 90-day increases as the CV value rises. Subgroup analysis showed an interaction between congestive heart failure and CV in 30-day and 60-day all-cause mortality, between age and CV in 60-day and 90-day all-cause mortality, and between chronic pulmonary disease and CV in 30-day all-cause mortality (<i>P</i> all &lt; 0.05).</p> Conclusion <p>The CV is associated with the all-cause mortality of AP patients in the ICU, especially when the CV value is between 0.25 and 0.45. When using CV, the effects of age, congestive heart failure, and chronic pulmonary disease should be considered.</p>