Recent prevalent use of three-dimensional image-guided brachytherapy (3D brachytherapy) has dramatically improved the treatment outcomes of cervical cancer. Inverse planning simulated annealing (IPSA) is one of the commonly used algorithms in 3D brachytherapy, but different conditions may affect the treatment plan quality. In this study, we compared HRCTV (high-risk clinical target volume) D90 (dose prescription) and HRCTV D95 D2cc (dose received by 2.0cc) of the rectum, bladder, and sigmoid in 30 patients with cervical cancer under four IPSA conditions. The HRCTV D90 (mean ± SD cGy) was 607.32 ± 37.86, 599.01 ± 23.62, 598.67 ± 13.07, and 596.45 ± 10.94 in four groups, respectively. The HRCTV D95 was 558.19 ± 38.51, 558.17 ± 25.72, 557.03 ± 16.12, and 555.26 ± 12.78, respectively. The sigmoid D2cc was 282.96 ± 44.84, 273.14 ± 60.69, 268.94 ± 62.32, and 292.69 ± 52.44. HRCTV D90, HRCTV D95, and sigmoid D2cc were not statistically different among the four groups (p > 0.05). However, the target fitness in group one, especially at the cervix, was poor. The rectum D2cc was 351.49 ± 32.90, 361.49 ± 28.09, 370.82 ± 24.44, and 375.33 ± 30.90. The rectum D2cc in group one was the lower than that in group three and group four ( p < 0.05 ). The bladder D2cc was 423.59 ± 31.39, 380.75 ± 37.25, 383.27 ± 32.55, and 385.22 ± 25.79. The bladder D2cc in group one was higher than the other groups ( p < 0.05 ). The maximum rectum limit dose (400cGy) is lower than the bladder (500cGy), and HRCTV is a whole in the IPSA algorithm; these result in the insufficiency or even absence of cervix dose that first need to meet in clinics. In conclusion, IPSA condition optimization can improve the quality of treatment plan in 3D brachytherapy and make it closer to clinical practice.