Cecal ligation and puncture (CLP) was used to cause severe sepsis in male Sprague-Dawley rats. There are four groups in this study: sham (n = 5), CLP (n = 10), end-tidal carbon dioxide (ETCO2) (n = 10), and mean arterial pressure (MAP) (n = 10). In ETCO2 group, fluid resuscitation (FR) began when ETCO2 at most 25 mmHg. In MAP group, FR began when MAP at most 100 mmHg. Electrocardiogram, aortic pressure, core temperature, and ETCO2 values were recorded at baseline, 2, 4, 6, 8, 10, and 12 h post-CLP. Lactate level, cardiac output (CO), perfused small vessel density (PSVD), and microvascular flow index (MFI) were assessed at the same time points as above. The results showed that MAP, CO, and ETCO2 gradually decreased after CLP. After FR, MAP, ETCO2, and CO in the ETCO2 group increased compared with the MAP group 12 h after CLP (all P < 0.05). Lactate level remains high in MAP group while decreasing in the ETCO2 group 8 h post-CLP. Both PSVD and MFI deteriorated after CLP in CLP group, though significantly improved in the ETCO2 group 8 h post-CLP. The average survival time in the ETCO2 group was significantly greater than MAP group (14.95 ± 3.90 h vs. 11.15 ± 1.76 h; t = 2.804, P = 0.012). Moreover, ETCO2 showed a negative correlation with lactic acid levels and a positive correlation with CO, PSVD, and MFI. In conclusion, ETCO2 can guide FR implement and improve outcomes of severe sepsis in CLP-inducted rat model. ETCO2 might be a potential index to guiding early FR in severe sepsis.