Measurements of a well-characterised standard sample can verify the performance of an instrument. Typically, small-angle neutron scattering instruments are used to investigate a wide range of samples and may often be used in a number of configurations. Appropriate standard samples are useful to test different aspects of the performance of hardware as well as that of the data reduction and analysis software. Measurements on a number of instruments with different intrinsic characteristics and designs in a round robin can not only better characterise the performance for a wider range of conditions but also, perhaps more importantly, reveal the limits of the current state of the art of small-angle scattering. The exercise, followed by detailed analysis, tests the limits of current understanding as well as uncovers often forgotten assumptions, simplifications and approximations that underpin the current practice of the technique. This paper describes measurements of polystyrene latex, radius 72 nm with a number of instruments. Scattering from monodisperse, uniform spherical particles is simple to calculate and displays sharp minima. Such data test the calibrations of intensity, wavelength and resolution as well as the detector response. Smoothing due to resolution, multiple scattering and polydispersity has been determined. Sources of uncertainty are often related to systematic deviations and calibrations rather than random counting errors. The study has prompted development of software to treat modest multiple scattering and to better model the instrument resolution. These measurements also allow checks of data reduction algorithms and have identified how they can be improved. The reproducibility and the reliability of instruments and the accuracy of parameters derived from the data are described.