Objective To analyze the characteristics of imported malaria in Changsha and explore the influencing factors and early warning indicators of severe malaria patients in combination with laboratory indicators. Methods A retrospective analysis was conducted on the case data of malaria patients hospitalized in the First Hospital of Changsha from 2012 to 2023. Differences between non-severe and severe malaria patients were compared. Bivariate correlation-Spearman correlation coefficient and logistic regression analysis were used to observe the correlation and influence degree between indicators and severe malaria. Indicators strongly related to prognosis were screened to identify independent impact factors for the progression of malaria patients to severe cases. Receiver operating characteristic (ROC) curve analysis was performed to assess the efficacy of individual and combined indicators in early prediction of severe malaria, and the area under the curve (AUC) was calculated. Results Case data from 127 malaria patients were collected, which including 38 severe malaria patients and 89 non-severe patients. Compared with non-severe malaria patients, severe malaria patients had longer durations from onset to hospitalization [4.00 (2.75, 6.00) d] and longer hospitalization time [7.00 (4.75, 12.00) d] (Z = -2.20, -3.05, P < 0.05, P < 0.01). More severe malaria patients [78.9% (30/38)] were presenting with occult blood in urine (χ² = 18.67, P < 0.01). They had lower red blood cell count [4.11 (3.46, 4.65) × 10¹²/L], platelet count [(53.50 ± 53.73) × 10⁹/L] and albumin level [(34.99 ± 5.36) g/L] (Z = -2.37, t = 6.72, t = 4.10; P < 0.05, P < 0.01, P < 0.01), but higher levels of alanine aminotransferase (ALT) [53.50 (38.63, 73.75) U/L], aspartate aminotransferase (AST) [50.85 (34.00, 97.05) U/L], lactate dehydrogenase (LDH) [449.50 (321.50, 625.75) U/L], total bilirubin [(74.18 ± 78.09) μmol/L] and procalcitonin (PCT) [2.95 (0.91, 19.47) ng/ml] (Z = -3.51, Z = -3.48, Z = -3.58, t = -3.91, Z = -4.63; all P < 0.01) than non-severe patients. Durations from onset to hospitalization, hospitalization time, urinary occult blood and levels of ALT, AST, total bilirubin, LDH, and PCT were positively correlated with severe malaria (r > 0, all P < 0.05), while red blood cell count, platelet count and albumin level were negatively correlated with severe malaria (r < 0, all P < 0.05). Among them, platelet count was a good indicator of malaria severity (|r| ≥ 0.5, P < 0.01). Urinary occult blood, AST, total bilirubin, LDH, serum creatinine, urea nitrogen and PCT were risk factors for severe malaria (OR > 1, P < 0.05), while red blood cell count, platelet count, hemoglobin and albumin level were protective factors against severe malaria (OR < 1, P < 0.05). Platelet count and total bilirubin level were independent impact factors for the progression of malaria patients to severe cases (P < 0.05), with AUCs of 0.867 and 0.769, respectively. The AUC for the combined detection of the both was 0.900, which was higher than that for single indicator detection (P < 0.01). Conclusion Platelet count and total bilirubin level can serve as independent impact factors for early warning of severe malaria. Combined detection is more helpful than single detection for early warning of severe malaria.