Objective To detect the effect of predictive value of blood paraquat (PQ) concentrations at different time intervals from ingestion to blood collection. Methods 284 patients with acute PQ poisoning from October 2010 to May 2023 from Cangzhou Central Hospital were enrolled in this retrospective study, and divided into survival and dead groups according to a 90-day prognosis. The logistic regression was used to analyze the predictive factors for prognosis of patients with acute PQ poisoning. The time intervals from ingestion to blood collection were four time periods: ≤4.0 h, 4.1~8.0 h, 8.1~12.0 h, and ≥12.1 h. The area under the curve (AUC), sensitivity, and specificity of blood PQ concentrations for patient's prognosis at each time interval were calculated by the receiver operating characteristic (ROC) curve. Results The 90-day survival rate was 45.8% (130/284). Compared with the dead group, the patients in survival group had younger age (Z=3.367, P<0.001), decreased white blood cells (Z=9.167, P<0.001), alanine aminotransferase (Z=2.237, P=0.025), blood creatinine (Z=9.373, P<0.001), blood amylase (Z=2.003, P=0.045), blood lactate (Z=10.116, P<0.001), and blood PQ concentration (Z=11.325, P<0.001), while increased blood potassium (Z=7.692, P<0.001), PaCO2(Z=7.884, P<0.001), and base excess (Z=10.506, P<0.001). Logistic regression analysis showed that age (OR=1.036, 95%CI: 1.009-1.064, P=0.008), white blood cells (OR=1.106, 95%CI: 1.004-1.219, P=0.042), blood creatinine (OR=1.030, 95%CI: 1.015-1.045, P<0.001), blood potassium (OR=0.318, 95%CI: 0.125-0.810, P=0.016), base excess (OR=0.737, 95%CI : 0.625-0.869, P<0.001), and blood PQ concentration (OR=2.388, 95%CI: 1.596-3.573, P<0.001) were independent predictive factors for prognosis. Blood PQ concentration was positively correlated with prognosis (r=0.358, P<0.001), with a mortality rate of 20.0% for blood PQ concentration≤1.0 μg/mL, 54.8% for 1.1~2.0 μg/mL, 64.3% for 2.1~3.0 μg/mL, 88.9% for 3.1~4.0 μg/mL, and 100% for concentration≥4.1 μg/mL. ROC curve analysis showed that the AUCs of blood PQ concentration were 0.922 (95%CI: 0.888~0.956), 0.912 (95%CI : 0.834~0.990), 0.813 (95%CI: 0.465~1.000), and 0.522 (95%CI: 0.207~0.837) for ≤4.0 h, 4.1~8.0 h, 8.1~12.0 h, and ≥12 1 h. Conclusion The predictive value of blood PQ concentrations is influenced by the time from ingestion to blood collection, the predictive value of PQ concentration within 4 h is good, and poor beyond 8 h