In 1989, 1990 and 1991 five-week-old female broilers were X-rayed in a computer tomograph (scanned), and then dissected. The broilers were scanned alive in two different positions; one cutting through the anterior end of the sternum, where the pectoral muscles are located, and the other cutting through the posterior end of the sternum, where the abdominal fat is placed. The broilers were dissected for breast cut and for abdominal fat. Regression equations with dissection results as dependent variables were used as prediction equations for a sample other than the experimental birds used to make the equations. The validation sample could be either from the same year as the sample from which the regression equation was made, or from different years. Correlations between dissection results and predicted results on samples from the same year were between 0.63 and 0.70 for abdominal fat and between' 0.54 and 0.76 for breast cut. When comparing samples from different years, however, the correlations considering abdominal fat were significantly reduced.