Lame and sound broilers, selected from commercial flocks, were trained to d
iscriminate between different coloured feeds, one of which contained carpro
fen. The two feeds were then offered simultaneously and the birds were allo
wed to select their own diet from the two feeds, in an initial study to ass
ess the most appropriate concentration of drug, the plasma concentrations o
f carprofen were linearly related to the birds' dietary intake. The walking
ability of lame birds was also significantly improved in a dose-dependent
manner and lame birds tended to consume more analgesic than sound birds. In
a second study, in which only one concentration of analgesic was used, lam
e birds selected significantly more drugged feed than sound birds, end that
as the severity of the lameness increased, lame birds consumed a significa
ntly higher proportion of the drugged feed.