Although high-volume online hemodiafiltration has been associated with higher clearance and lower pre-dialysis concentration of middle molecular weight toxins compared to hemodialysis, its effect on protein-bound uremic toxins has shown inconclusive results. In this study, we investigated whether hemodiafiltration impacts pre-dialysis plasma levels of the toxins indoxyl sulfate, p-cresyl sulfate, and indole-3-acetic acid compared to high-flux hemodialysis. This is a post-hoc analysis of the multicenter, randomized controlled trial HDFit (ClinicalTrials.gov: NCT02787161). Uremic toxins were determined by high performance liquid chromatography at baseline, 3, and 6months. Mean differences in monthly changes of pre-dialysis uremic toxin concentrations between hemodiafiltration and high-flux hemodialysis were analyzed using linear mixed-effect models. One hundred ninety-three patients (mean age 53years old, 71% males) were analyzed. There were no differences between groups regarding clinical and biochemical characteristics at baseline or duration of dialysis session and blood flows throughout the follow-up. Mean differences in rates of change (μM/month, [confidence interval CI]) in high-flux hemodialysis vs. hemodiafiltration were 2.4 [0.3 to 4.56], 3.94 [- 1.54 to 9.41] and 0.06 [- 0.6 to 0.5] for indoxyl sulfate, p-cresyl sulfate and indole-3-acetic acid, respectively. In the exploratory analysis, these differences in high-flux hemodialysis vs. hemodiafiltration subgroup with convective volume > 27.5L were 2.86 [0.43 to 5.28], 7.43 [0.7 to 14.16] and - 0.19 [- 0.88 to 0.50]. These exploratory findings suggest that hemodiafiltration is more effective in reducing indoxyl sulfate as compared to standard high-flux hemodialysis, and also that this effect was extended to p-cresyl sulfate in patients achieving higher convective volumes.