This study investigates how wind shear and momentum fluxes in the surface- and boundary layer vary across wind and cloud regimes. We use a nine-year-long data set from the Cabauw tower of the Ruisdael Observatory (NL) complemented by (8.2 x 8.2 km^2) daily LES hindcasts. An automated algorithm classifies observed and simulated days into different cloud regimes: 1) clear-sky days, 2) days with convective clouds (cumulus) rooted in the surface layer, with three ranges of cloud cover, and 3) days with clouds not rooted near the surface. Categorized days in observations and LES do not fully match, with a tendency of the LES to develop convective clouds on clear-sky days and less frequently produce non-rooted clouds, whose scales are far larger than the LES domain. Even so, the climatology and diurnal cycle of winds are for each regime very similar in LES and observations, strengthening our confidence in LES’ skill to reproduce certain clouds for an atmospheric state. Wind shear is smallest in clear-sky and cumulus regimes with limited cloud cover (CLCC), which also have the weakest 200 m wind speed and largest surface buoyancy flux. They have notably larger cross-wind fluxes, although along-wind momentum flux profiles are similar across all regimes. Cloudy days have larger momentum fluxes distributed over deeper layers, sustaining up to 20% of the surface flux value at cloud base. Compared to clear-sky, the CLCC regimes have stronger updrafts and deeper mixed-layers. At similar atmospheric stability, surface friction is larger and underestimated by Monin-Obukhov Similarity Theory.