Objective To explore the relevant risk factors that affect the early treatment effect of severe coronavirus disease 2019 (COVID-19). Methods A retrospective analysis was performed on the data of 71 severe COVID-19 patients who were admitted to Hankou Hospital, Wuhan, Hubei from Jan. to Feb. 2020 with positive in nucleic acid test of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The early vital signs, clinical manifestations, resting oxygen saturation, blood routine, liver function, blood biochemistry, electrolyte, high-sensitivity troponin, coagulation function, pro-brain natriuretic peptide, lung CT manifestations and sequential organ failure assessment (SOFA) at admission, as well as treatment regimens and early treatment outcomes were collected. According to the treatment effect after 2 weeks, the patients were divided into clinical remission group and clinical deterioration group. The differences of each index between the two groups were compared, and the risk factors affecting the early treatment effect were analyzed. Results Forty-three patients in the clinical remission group turned into non-severe cases after treatment. Among the 28 patients in the clinical deterioration group, 10 died, 15 had disease progression to critically ill status, and three had no significant improvement after treatment. The patients in both groups had a history of staying in Wuhan before the onset of the disease. The proportion of smokers in the clinical deterioration group was significantly higher than that in the clinical remission group (75.0% [21/28] vs 46.5% [20/43], P=0.033). Compared with the patients in the clinical remission group, the patients in the clinical deterioration group were more likely to have different degrees of reductions in lymphocyte counts (0.80 [0.70, 0.90]×10⁹/L vs 0.70 [0.60, 0.70]×10⁹/L), platelet counts (222 [174, 310]×10⁹/L vs 193 [152, 232]×10⁹/L) and fibrinogen level (4.22 [3.71, 4.80] g/L vs 3.81 [2.96, 4.38] g/L) (all P<0.05). There were no significant differences in common clinical symptoms, vital signs, resting oxygen saturation, other laboratory indicators, SOFA score, or treatment regimens between the two groups (all P>0.05). Multivariate logistic regression analysis revealed that smoking (odds ratio [OR]=4.88, 95% confidence interval [CI] 1.33-25.00, P=0.020), white blood cell (WBC) count ≤3.5×10⁹/L (OR=10.00, 95% CI 1.47-100.00, P=0.008), and lymphocyte count <0.1×10⁹/L (OR=16.67, 95% CI 3.33-100.00, P<0.001), were the independent risk factors affecting the early treatment effect of severe COVID-19 patients. Conclusion The severe COVID-19 patients with smoking history, WBC count ≤3.5×10⁹/L or lymphocyte count<0.1×10⁹/L have a higher risk of poor early treatment, and more attention should be paid in clinical diagnosis and treatment of these patients to improve the prognosis.