A method is described for the sedimentation velocity analysis of solutions composed of macromolecular solutes of widely disparate size. In sedimentation velocity experiments, usually a single rotor speed is chosen for the entire run, and consequently, the range of observable sedimentation coefficients can be severely limited. This limitation can be removed if the speed is varied during the run, starting with a relatively low speed so that the largest particles can be easily observed. The speed is increased during the run until full speed is attained and the run continued at full speed until the smallest species of interest have cleared the solution. The method, called wide distribution analysis, is based on the method developed originally by Yphantis and co-workers (Proc. Natl. Acad. Sci. USA 78 (1981) 1431) and on the time derivative method of Stafford (Anal. Biochem. 203 (1992) 295), essentially eliminating both the time-independent and radially-independent noise thereby improving the precision, especially for interference optics. An algorithm for analysis of data from both absorbance and interference optics and experimental protocols compatible with the Beckman XL-I Analytical Ultracentrifuge are presented. With these protocols an extremely wide range of sedimentation coefficients from approximately 1.0 to 250000 S can be accommodated in a single multi-speed run.