TYPE: Case Report TOPIC: Critical Care PURPOSE: To evaluate the effects of platelet count and its variation trend on patient with infectious diseases in intensive care units (ICUs). METHODS: This retrospective cohort study extracted 4,251 critically ill adult patients with infectious diseases from the eICU Collaborative Research Database, whose platelet counts were measured daily during the first 7 days after admission. The variation trends in platelet counts were compared between the survivors and non-survivors. Logistic regression analyses were performed to analyze the relationship between nadir platelet count and ICU mortality. RESULTS: In the survivors, platelet counts decreased in the first days after admission, reached a nadir on day 3, and then returned and continued to rise above the admission value. In non-survivors, the platelet counts decreased after admission, without a subsequent upturn. We defined three subgroups according to the nadir platelet counts: ≤50, 50–130, and ≥130×109/L, corresponding to low, intermediate, and high ICU mortality. A decreased platelet count was associated with increased ICU mortality (intermediate vs. low: 1.651 [1.266–2.154]; high vs. low: 3.785 [2.722–5.263]). CONCLUSIONS: During the first 7 days, platelet counts decreased after ICU admission, while increased subsequently in the survivors but not in the non-survivors. ICU mortality risk increased as nadir platelet count decreased below to 130×109/L, and further boosted when the nadir platelet count level reached below 50×109/L. CLINICAL IMPLICATIONS: Combining platelet count and its variation trend may better reveal the prognosis of critically ill patients with infectious diseases. DISCLOSURE: Nothing to declare. KEYWORD: Platelet count