Experiments in England from 1970 to 1973 investigated the possibility of extending the growing season and increasing yields of bolting resistant sugar beet by sowing in autumn rather than in spring. September-sown plants grew rapidly in spring and achieved a complete leaf cover by mid-June by which time they had intercepted 40 % of radiation incident since mid-April compared with 28 and 9% for crops sown on 13 October and 13 April respectively. By early June autumn-sown plants were 20 fold heavier than spring sown but this advantage was progressively lost with the onset of bolting. Root growth slowed as assimilates were diverted to stem and inflorescence growth. Delaying autumn sowing from late September until mid-October delayed and reduced the frequency of bolting, probably because of a shortening of the period when plants were sufficiently large to respond to cold. Cutting down bolting inflorescences at frequent intervals improved root growth of both bolters and non-bolters. However, yields from autumn sowing never exceeded those from spring sowing, the best yield from an autumn-sown crop was 6·1 t/ha of sugar which was similar to that of the spring-sown crop. To try to prevent bolting of overwintered plants the growth regulator ethephon was applied at 10000 mg a.i./l water sufficient to wet the foliage in mid-April. Growth was severely checked and 55 % of plants died. Of the remainder 25–30% fewer than in the untreated crop bolted but yields were 26–42 % less. Less concentrated doses of ethephon did not affect bolting.