Two approaches have been taken to the modeling of poultry demand in U.S. meat demand studies. One has been to ignore turkey, and estimate demands for beef, pork, and chicken. The second has been to include turkey by combining it with chicken, and estimating demands for beef, pork, and poultry. The validity of these two approaches is examined using quarterly U.S. time-series data from 1980-96. The results indicate that either approach to the modeling of poultry demand is appropriate.