Studies in the 1970s showed that sugar beet partitioned into root, crown and scalp had large concentrations of impurities in the scalp, the root portion had the best processing quality and the crown was intermediate. Since then the processing quality of beet has improved. This paper reports similar studies on today's varieties, in an experiment at IACR-Broom's Barn over the three years 1993±1995, and extended in 1996 to the most important soil types used for sugar beet production in England. The crown is now a smaller proportion of the total beet weight (70 compared to 130 g kg À1 ) but the distribution of the concentrations of the quality parameters (sugar and the impurities potassium, sodium, amino nitrogen, raf®nose, betaine and invert sugars) between the beet portions has not changed. This distribution was not signi®cantly affected by year, variety or beet size but was affected by soil type. Nitrogen-rich soils produced beet with heavier crowns and a smaller differential in impurity concentration between the root and the crown. The scalp material had a variable sugar concentration, but it always contained large concentrations of impurities, especially invert sugar. The equation of Pollach et al (Zucker industrie 116: 689±700 (1991)) which includes invert sugar as a variable, was applied to the data. It indicated that including all the crowns in the factory process with the roots would increase white sugar yield by approximately 6%. The increase due to inclusion of the scalp was much smaller (2%), and would probably be expensive to achieve.