Background To compare the efficacy of Ex-PRESS implantation versus trabeculectomy combined with phacoemulsification. Methods A retrospective 12-month study on patients with coincident primary open-angle glaucoma (POAG) and cataract. The patients underwent combined phacoemulsification and Ex-PRESS implant (Phaco-ExPRESS, n = 35) or phacotrabeculectomy (Phaco-Trab, n = 35). The morphological structures of the filtering bleb were examined by slit-lamp, anterior segment optical coherence tomography (AS-OCT) and in vivo confocal microscopy (IVCM). Complete success was defined as postoperative intraocular pressure (IOP) < 18 mmHg without the use of anti-glaucoma medication. Qualified success was defined as postoperative IOP < 18 mmHg with or without anti-glaucoma medications. The data were collected preoperatively and postoperatively at 2 weeks, 1 month, 3 months, 6 months, and 12 months. Results No significant difference in the variables such as age, IOP and perimetry was found between the groups of Phaco-ExPRESS and Phaco-Trab. At the one-year postoperative visit for filtering blebs, Phaco-ExPRESS increased the mean area of epithelial microcysts significantly from 0.10 ± 0.05 to 0.20 ± 0.09 μm2 per μm2, while Phaco-Trab decreased the mean area significantly from 0.08 ± 0.04 to 0.04 ± 0.06 μm2 per μm2. Notably, the hyperreflective dots detected by IVCM decreased by 84.9% in Phaco-ExPRESS but increased by 36.3% in Phaco-Trab. The hyperreflective dots were further identified as neutrophil- and monocyte-like cells. The number of these cells were negatively correlated with the microcysts area (r =  − 0.7, P < 0.01) but positively associated with the grade of connective tissue (r = 0.5, P < 0.01). By creating different microstructural changes in the filtering blebs, Phaco-ExPRESS produced a higher complete success rate (84.9% vs. 41.2%, P < 0.01) and significant decrease in the number of anti-glaucoma medications (P < 0.01) when compared with those in Phaco-Trab. However, the qualified success showed no significant difference between the two groups (100.0% vs. 91.2%, P = 0.24). Conclusions At the one-year follow-up, Phaco-ExPRESS generated better filtering bleb with larger area of microcysts, looser connective tissues, and less inflammation than that of Phaco-Trab, providing adequate IOP control and less IOP-lowering medications. These findings indicate that Phaco-ExPRESS could be more preferred than Phaco-Trab for the treatment of patients with coincident POAG and cataract.