AbstractIn a systematic study of the elution column fractionation of a well characterized branched polyethylene sample it was found that one of the most important factors contributing to the efficiency was the composition selected for the p‐xylene‐butyl cellosolve eluting mixture. With a mixture representing the critical solvent composition at 126°C., as determined by cloud point measurements, and deposition to an amorphous phase on a fine sand support, viscosity–distribution curves were obtained which were highly reproducible and completely free of any reversal. No further improvement was gained by using an elution temperature other than 126°C. or by the substitution of tetralin or mesitylene for p‐xylene. Precipitation down a temperature gradient to produce a physical separation of species provided enhanced fractionation but is of doubtful practical interest. Fractionation was readily scaled up from 2g. to 4 or 6g., merely by increasing the mixing vessel size from 500 cc. to 1000 cc. There was no loss of efficiency, and fractions with M̄w = 8 × 106, more than twentyfold higher than the sample value, were obtained despite eluted polymer concentrations exceeding 0.2%. There is evidence that many of the expected differences in the fractionation behavior of linear and branched polyethylene are erased when a solvent mixture is used which has been adjusted to the appropriate critical composition.