Dynamic light scattering (DLS) is often used to monitor aggregation in protein solutions. Here, we explore the veracity of the aggregate sizes, size distribution widths, concentrations, and lifetime resulting from DLS. We use as an example a solution of the protein lysozyme in which dense liquid clusters of radius about 100 nm reproducibly exist. We compare the results of DLS to those of brownian microscopy. We show that because of the sixth power dependence of the scattered light intensity on the size of the scatterers, DLS overestimates the mean size of the clusters. The factor of overestimation depends on the shape of the size distribution and is ∼1.6 × in the studied solution. The related underestimate of the cluster concentration is ∼10 ×. The CONTIN algorithm, often employed to process DLS data, may, in some instances, produce non-physical results. We put forth an alternative method to determine the aggregates' sizes, concentrations, and volume fractions. We show that DLS yields a reliable width of the cluster size distribution only if the cluster concentration is above 10(9) cm(-3) and their volume fraction is above 10(-6). DLS yields a lower bound of the cluster lifetime, which may be orders of magnitude lower than the real one.