To investigate the value of procalcitonin (PCT) on predicting the severity and prognosis in patients with early acute respiratory distress syndrome (ARDS). A prospective observation study was conducted. A total of 113 patients with ARDS undergoing mechanical ventilation admitted to intensive care unit (ICU) of Affiliated People's Hospital of Jiangsu University from October 2012 to April 2016 were enrolled. Based on oxygenation index (PaO2/FiO2), the patients were classified into mild, moderate, and severe groups according to Berlin Definition. Twenty-five healthy volunteers were served as controls. Demographics, acute physiology and chronic health evaluation II (APACHE II) score, and Murray lung injury score were recorded. Within 24 hours after diagnosis of ARDS, the serum levels of PCT and C-reactive protein (CRP) were determined by enzyme-linked fluorescence analysis (ELFA) and immune turbidimetric method, respectively. The patients were also divided into survival and non-survival groups according to clinical outcome within 28-day follow-up, and the clinical data were compared between the two groups. Spearman rank correlation was applied to determine the correlation between variables. The predictive value of the parameters on 28-day mortality was evaluated with receiver operating characteristic curve (ROC). Kaplan-Meier survival curve analysis was used to compare different PCT levels of patients with 28-day cumulative survival rate. After excluding patients who did not meet the inclusion criteria and loss to follow-up, the final 89 patients were enrolled in the analysis. Among 89 ARDS patients analyzed, 27 of them were mild, 34 moderate, and 28 severe ARDS. No significant differences were found in age and gender between ARDS and healthy control groups. Infection and trauma were the most common etiology of ARDS (55.1% and 34.8%, respectively). Compared with healthy control group, both CRP and PCT in serum of ARDS group were higher [CRP (mg/L): 146.32 (111.31, 168.49) vs. 6.08 (4.47, 7.89), PCT (μg/L): 3.46 (1.98, 5.56) vs. 0.02 (0.01, 0.04), both P < 0.01], and the two showed sustained upward trends with the ARDS course of disease. Compared with mild group, severe group had significantly higher APACHE II and Murray scores. Spearman rank correlation analysis showed that both serum PCT and CRP in patients with ARDS was correlated well with APACHE II score (r values were 0.669 and 0.601, respectively, both P < 0.001), while PCT was weakly but positively correlated with Murray score (r = 0.294, P = 0.005), but not the case of CRP (r = 0.203, P = 0.052). APACHE II score and serum PCT in non-survival group (n = 38) were significantly higher than those of the survival group [n = 51; APACHE II score: 26.00 (23.00, 28.50) vs. 21.00 (17.00, 25.00), PCT (μg/L): 6.38 (2.82, 9.49) vs. 3.09 (1.08, 3.57), both P < 0.01], but Murray score and CRP level were similar between survivors and non-survivors. The areas under ROC curve (AUC) of APACHE II score and PCT for predicting 28-day mortality were 0.781 and 0.793, respectively, which were better than those of AUC of Murray score and CRP (0.606 and 0.561, respectively, all P < 0.05). The AUC of APACHE II score combined with PCT was significantly higher than that of PCT (0.859 vs. 0.793, P = 0.048) or APACHE II score alone (0.859 vs. 0.781, P = 0.038). Using a PCT cut-off value of > 4.35 μg/L for predicting 28-day mortality, the sensitivity and specificity was 92.2% and 63.2%, respectively, and the positive and negative likelihood ratios were 2.50 and 0.12 respectively. Kaplan-Meier survival curve analysis indicated that the patients whose PCT more than 4.35 μg/L, had lower 28-day cummulative survival rate as compared with those with PCT ≤ 4.35 μg/L (log-rank test: χ2 = 5.013, P = 0.025). The elevated serum PCT level in patients with ARDS seems to be correlated well with the severity of lung injury, and appears to be a useful prognostic marker of outcome in the early phases of ARDS.