Between 1997 and 1999 three weed control trials were carried out with sugar beet (Beta vulgaris L.) resistant to glyphosate. Glyphosate was applied at a total dose of 1620, 2160, 3240 and 4320 g active ingredient (a.i.)/ha over a three-spray programme and at 1440, 2160 and 4320 g a.i./ha over a two-spray programme. These treatments were compared with standard (1N) and double dose (2N) three-spray sugar-beet herbicide programmes. In each year, the lowest glyphosate dose, applied as a three-spray programme, gave marginally better overall control of weeds than the standard herbicide programme. In all trials the weed control improved after each application. However, no significant differences in weed control were observed after the second glyphosate application; the three-spray glyphosate programmes gave marginally, but not significantly, better weed control than the two-spray programme in all seasons. At all doses more than one glyphosate application was required only in respect of Falopia convolvulus and Lamium purpureum for total weed control. In 1997 the herbicide-tolerant sugar-beet variety sown was not completely tolerant to glyphosate and a reduction in plant numbers was recorded after the initial glyphosate application. The variety used in the two subsequent years was fully tolerant and no plant loss was observed, even at the highest glyphosate dose. At harvest, most of the 1998 root yields in the glyphosate treatments were significantly higher than the yields from the standard (1N) herbicide programmes. This was not the case in 1999. Crop vigour was not affected by any of the treatments in 1997 but in 1998 and 1999 the two standard herbicides reduced crop vigour by 10 and 20% and 10 and 15%, respectively.
