Experiments at Sutton Bonington in 1973 and 1974 compared the growth and yield of sugar beet which was handweeded at different stages with that where weed control was entirely by pre- and post-emergence herbicides, used alone and in various combinations. Data from the series of handweeding treatments enabled the crop's minimum weed control requirements to be specified. They indicated that an ideal herbicide would kill all the weeds present when applied at the six-leaf stage, persist in the soil to prevent weed infestation for 4 weeks, but not check the crop to the extent that recovery was incomplete. That the crop was capable of recovering from a degree of early check was clearly evident from the handweeded treatments; it proved possible for crops whose weights had been depressed by 25–30 % when first weeded to recover completely by harvest.

The performance of all post-emergence treatments fell short of these requirements. When the crop was at the six-leaf stage, the earliest emerging weeds were too far advanced to be killed by the herbicide. By themselves, pre-emergence herbicides gave weed control equivalent to only 2 or 3 weeks of handweeding but they did pave the way for more effective use of post-emergence treatments by restricting the spectrum of weed sizes present and sensitizing those plants which did establish. The most effective herbicide regime used in the experiments was the sequence of chloridazon applied pre-emergence followed by two post-emergence applications of phenmedipham. However, phenmedipham clearly checked the crop. Two weeks after application crop dry weight was usually depressed by one-third. Provided weed control was adequate, recovery was usually complete by the end of the season.