This study aimed to assess refractive astigmatism, in phaco-canaloplasty (PC) vs phaco-non-penetrating deep sclerectomy (PDS) in a randomized, prospective study within 24 months. Patients were randomized pre-operatively, 37 underwent PC and 38 PDS. The following data was collected: BCVA, IOP, number of antiglaucoma medications, refraction with autokeratorefractometry. The assessment of astigmatism was simple arithmetic and vector analysis (calculations included cylinder with axis in form of centroids) and included double angle plots and cumulative refractive astigmatism graphs. Pre-operative mean BCVA in PC was 0.40 ± 0.43 and was comparable to BCVA in PDS 0.30 ± 0.32logMAR (P = 0.314). In the sixth month follow-up, mean BCVA showed no difference (P = 0.708) and was 0.07 ± 0.13 and 0.05 ± 0.11, respectively. However, 2 years after the intervention mean BCVA was better in PC 0.05 ± 0.12 than in PDS 0.12 ± 0.23 and it was statistically significant (P = 0.039). Mean astigmatism in PC at baseline was 1.13 ± 0.73Dcyl, at 6 months it was 1.09 ± 0.61 and at 2 years 1.17 ± 0.51. In PDS at baseline 1.35 ± 0.91 at 6 months 1.24 ± 0.86 and at 2 years 1.24 ± 0.82. There were no differences between the groups in mean astigmatism throughout the study. Centroids (mean of a cylinder with axis) in PC were pre-operatively 0.79D@172˚ ± 1.10Dcyl, at 6 months 0.75D@166˚ ± 1.01 and at 24-months 0.64D@164˚ ± 1.11 and in PDS pre-operatively 0.28D@10˚ ± 1.63D at 6 months 0.26D@11˚ ± 1.5 and at 24-months 0.47D@20˚ ± 1.43. The direction of mean astigmatism was against the rule in all analyzed time points. The mean baseline IOP in PC was 19.4 ± 5.8 mmHg and 19.7 ± 5.4 mmHg in PDS(P = 0.639). From the 6-month IOP was lower in PC, at 24-months it was 13.8 ± 3.3 mmHg in PC and 15.1 ± 2.9 mmHg in PDS(P = 0.048). In both groups preoperatively patients used median(Me) of three antiglaucoma medications(P = 0.197), at 24-months in PC mean 0.5 ± 0.9 Me = 0.0 and 1.1 ± 1.2 Me = 1.0 in PDS(P = 0.058). Both surgeries in mid-term observation are safe and effective. They do not generate vision-threatening astigmatism and do not even change the preoperative direction of mean astigmatism. Refractive astigmatism is stable throughout the observation.