Failure to control Bremia lactucae (lettuce downy mildew) with metalaxyl in an intensive lettuce-producing region of Lancashire at the end of 1983 was shown to be due to the occurrence of a high level of resistance to this fungicide (isolates capable of growth at < 100 μg/ml metalaxyl). During most of 1984, metalaxyl-resistant isolates were obtained from numerous sites but all within a 20-km radius of the initial outbreak. Thereafter, at the end of 1984 and during 1985, metalaxyl-resistant isolates were recovered from most major lettuce-producing regions in the UK with protected crops more affected than field crops. AH metalaxyl-resistant isolates tested were identical in their response to fungicide, sexual compatibility type (B2) and virulence phenotype, probably representing a clone from a single origin. The resistant pathotype was virulent on resistance factors R 1-10 and 12-15 but lacked virulence for R 11 and 16-18. This was also the most common virulence phenotype among sensitive isolates collected at the same time. Cross-resistance to other phenylamide fungicides was demonstrated but isolates resistant and sensitive to phenylamide showed a similar response to the unrelated systemic fungicides propamocarb and fosetyl-Al. An F1 sexual progeny isolate from a cross between a phenylamide-sensitive and a phenylamide-resistant isolate (presumed heterozygous at the locus or loci regulating response to phenylamide fungicides) exhibited an intermediate response to phenylamide fungicides. No isolates of this type were obtained from the field. At the high concentrations affecting spore germination, phenylamide fungicides exhibited lower activity against a resistant isolate compared with a sensitive isolate. The findings are discussed in relation to future control strategies, the population biology of the fungus and possible directions for lettuce breeding programmes.