The National Scrapie Plan (NSP) was launched in Great Britain in 2001, with the aim of eventually eradicating scrapie, a small ruminant transmissible spongiform encephalopathy, from the national sheep flock. Specifically, a selective breeding programme, the Ram Genotyping Scheme, was devised enabling pedigree ram breeders to reduce the number of scrapie-susceptible genotypes from their flocks. The effect of large-scale manipulation of PrP genotypes on commercially important traits within the sheep industry is, however, unknown. We have therefore examined production traits in a total of 43 968 lambs from 32 pedigree breeders across three British hill breeds, comprising 8163 North Country Cheviot (Hill), 21 366 Scottish Blackface and 14 439 Welsh Mountain lambs. Traits examined included: weights at birth, 8 and 20 weeks; ultrasonic fat and muscle depth, and average daily weight gain from 8 to 20 weeks. Linear mixed models were fitted for each trait, including animal (direct) genetic effects and up to three maternal effects. Potential associations with the PrP gene were assessed by fitting either PrP genotype or number of copies of individual alleles as fixed effects. A number of breed-specific significant associations between production traits and the PrP gene were found, but no consistent significant effects were detected across the three breeds. Breed-specific effects were as follows: (i) 0.37 kg higher birth weights (BWTs) in AHQ homozygous North Country Cheviot (Hill) lambs (P < 0.01); (ii) 0.16 kg higher BWTs in ARR homozygous Scottish Blackface lambs (P < 0.05); (iii) 0.5 kg higher 8-week weights in VRQ heterozygous Scottish Blackface lambs (P < 0.01); (iv) a 0.72 kg decrease in scan weight associated with homozygous ARR Welsh Mountain lambs (P < 0.01); (v) 0.51 mm higher ultrasonic muscle depths in AHQ homozygous Welsh Mountain lambs (P < 0.01); (vi) 0.48 mm lower ultrasonic muscle depths in Welsh Mountain lambs carrying one or more copies of the ARR allele (P < 0.05) and (vii) 0.2 mm higher ultrasonic fat depths in heterozygous VRQ Welsh Mountain lambs (P < 0.05). The use of a Bonferroni correction to define appropriate significance thresholds across the three datasets, which account for the large number of independent comparisons made, resulted in breed-specific comparisons, with P < 0.01 becoming significant at P∼0.05, and the remaining breed-specific comparisons no longer being significant. The absence of a common effect across the three breeds suggests that any true association found may be due to breed-specific alleles of neighbouring genes in linkage disequilibrium with the PrP locus.