Objective To evaluate the accuracy of invasive systolic pressure variation (SPV) in monitoring the volume responsiveness of patients under pneumoperitoneum. Methods A total of sixty patients, aged 50‒70 years, American Society of Anesthesiologists (ASA) Ⅱ, with body mass index (BMI) of 19‒25 kg/m2, who were scheduled for laparoscopic radical resection of gastric cancer, were enrolled. Artificial pneumoperitoneum was established after endotracheal intubation under general anesthesia. Then, 3 min later, the volume loading test was performed, while 6% hydroxyethyl starch 130/0.4 injection was infused at 7 ml/kg over 15 min. Heart rate, mean arterial pressure (MAP), cardiac output (CO), cardiac index (CI), stroke volume index (SVI) and stroke volume variability (SVV) were recorded before pneumoperitoneum (T0), 3 min after pneumoperitoneum (T1), 3 min after the volume loading test (T2) and 2 h after surgery (T3). SPV was calculated after the title of invasive arterial pressure was changed. The patients were divided into two groups according to the increased percentage in SVI after the volume loading test (∆SVI): a positive volume responsiveness group (group R, n=29, ∆SVI≥10%), and a negative volume responsiveness group (group N, n=31, ∆SVI<10%). The receiver operator characteristic (ROC) curves of SPV and SVV were plotted, and the area under the receiver operator characteristic curve (AUC) and 95% confidence interval (CI) were calculated to determine the accuracy and diagnostic thresholds of SPV and SVV in monitoring the volume responsiveness of patients under pneumoperitoneum. Results Compared with those at T0, both groups presented increases in heart rate, MAP, SVI, CO and CI at T1 (P<0.05), without statistical differences in SPV and SVV (P>0.05). Compared with those at T1, both groups presented decreases in SPV and SVV at T2 (P<0.05), while increased SVI, CO and CI were found in group R (P<0.05). Compared with those at T2, all the indexes at T3 showed no statistical differences (P>0.05). Compared with group N, group R produced reduced SVI, and increased SPV and SVV at T0; increased SPV and SVV at T1; and increased SVI, CO and CI at T2 (P<0.05), without significant differences in other indexes (P>0.05). The AUC and 95%CI of SPV and SVV were 0.88 (0.77‒0.98) and 0.93 (0.87‒1.00), respectively. When SPV=6.5% was set as the cutoff value to monitor volume responsiveness, the sensitivity was 89.7% and the specificity was 87.1%. When SVV=10.5% was set as the cutoff value to monitor volume responsiveness, the sensitivity was 93.1% and the specificity was 80.6%. Conclusions SPV can be used to monitor the volume changes of patients under pneumoperitoneum.