Six small-angle X-ray cameras with block collimation systems were simulated, namely the original Kratky camera, a high-flux version of the Kratky camera, a SAXSess (Anton Parr) camera with a focusing mirror in a linear collimation setup and in a pin-hole setup, as well as a similar camera with a parallelizing mirror in a linear and a pin-hole setup. Their performance was examined using Monte Carlo ray-tracing. The Kratky and the SAXSess camera gave resolutions of 64–65 nm, the high-flux Kratky camera gave a resolution of 44 nm, and the camera with parallelizing mirror gave a resolution of 32 nm. The flux of the camera with parallelizing mirror was 1.47 times higher than for the SAXSess camera, and 18.6 times the flux of the Kratky camera. On changing the alignment, the camera with parallelizing mirror exhibited the best performance up to a resolution of 44 nm; the SAXSess camera was better for higher resolutions. Experimental flux measurements agree if no collimation system is added. Measurements of beam profiles and flux including collimation systems show only qualitative agreement because of user-dependent factors during alignment.