We report a method to determine the particle size distribution of small colloidal silica spheres via analytical ultracentrifugation and show that the average particle size, variance, standard deviation, and relative polydispersity can be obtained from a single sedimentation velocity (SV) analytical ultracentrifugation (AUC) experiment. The particle size distribution (psd) from the enhanced van HoldeWeischet (vHW) analysis accounts for the dynamic light scattering results quite well. In addition, the vHW psd equals the psd from a continuous distribution of sedimentation coefficients analysis where whole sedimentation velocity boundaries are fitted. The SV AUC interference optical data also yield the specific particle volume such that distributions of sedimentation coefficients for colloidal spheres can be converted directly to particle size distributions. Our results show that SV AUC experiments may yield a quantitative particle size distribution without a priori knowledge of the particle size and the shape of the size distribution.Synthetic macromolecules as well as colloids are always to a certain degree heterogeneous with respect to size and shape.For the size and shape determination, as well as the overall particle size distribution (psd), several techniques are available 1-3 such as transmission electron microscopy (TEM, or variations on this technique such as scanning electron microscopy and cryo-TEM) and atomic force microscopy (AFM). These microscopic techniques, except cryo-TEM, 4,5 do not image in situ. Drying of TEM samples may lead to particle shrinkage and distortion of particle structures, including aggregation due to capillary forces. Moreover, the electron beam may seriously damage the colloids via the melting and sintering of nanoparticles, as is the case for the small silica particles that are subject of the present study. Even if TEM images are fairly representative for the colloids in solution, many counts are needed for reliable statistics. For example, in the case of a polydispersity around 30%, typically at least 1000 counts are needed for a representative size distribution (see the case studies for various colloids in ref 1).Techniques that sample colloids in situ are static light scattering (SLS) and dynamic light scattering 6 (DLS). For sufficiently narrow size distributions, SLS data in the Guinier region can be analyzed with a momentum-expansion, to obtain an effective radius that is independent of the detailed shape of the distribution function. 1 For a broad size distribution, however, it is hardly possible to extract in any a priori fashion reliable distribution parameters from static light scattering profiles. DLS may be conducted on polydisperse colloids for which the particle shape is known, to obtain an apparent average particle size and standard deviation. The shape of the particle size distribution from DLS, however, may be highly questionable because of the ill-conditioned inversion problem encountered in DLS. 6 Since the scattered light intensity scales with the particle vo...