Abstract In reaction engineering, it is often that overcoming transport limitations improves reactor performance. This requires detailed analyses of transport phenomena in the catalytic beds. Nuclear magnetic resonance velocimetry (MRV) measurements have been utilized for analyzing mass transport of gas flows within opaque monoliths. Comparisons to full-field computational fluid dynamics (CFD) simulations, however, show significant deviations. In this study, polyethylene glycol (PEG) and 3D-printed monoliths including one open-cell foam (OCF) and one honeycomb were used to demonstrate that both operating fluid and monolith morphology influence the achievable signal-to-noise ratio and resolution of NMR data. The velocity profiles measured by MRV in OCF agreed well with full-field CFD simulations with ± 5% deviation. In addition, the similarity between the simulated and experimental velocity fields was quantified by the similarity index, which is 1 for identical images. A mean value of 0.83 was determined for a 10 PPI OCF. Thus, using PEG as the operating fluid and a 10 PPI OCF allows to improve both spatial resolution by 34% and the quality of agreement by 13 percentage points compared to the published results of gas velocimetry within 20 PPI OCF. We further identified and quantified possible sources of deviation between CFD and MRV velocity fields. By limiting our analysis to velocities higher than 45% of the maximum velocity, we could achieve similarity indices of 0.95–0.99.