A selection experiment with two lines of White Leghorns originating from a common base population was carried out over 5 generations with the aims of maintaining an unchanged egg weight, reducing age at first egg and reducing adult body weight. Each line consisted of 14 male and 42 female breeders. 2. Males were mass selected for low body weight at 20 weeks of age. To compensate for the expected correlated loss in egg weight, hens were selected according to an index which counteracted this undesirable change while also reducing age at first egg and reducing body weight. 3. An index value was calculated for each individual hen from average egg weight, age at first egg and body weight at 20 weeks. Index weights had to be calculated for each generation and line in accordance with the expected change in egg weight due to male selection on body weight. 4. For control matings hens with an index near the population average were mated either to males with body weight near the population average (control C1) or to the selected males within lines (control C2). 5. Expected and observed total responses agreed well for all traits in line 1 and for body weight in line 2. 6. Phenotypic variances and covariances showed little change during the experiment. However, genetic variances and covariances estimated at the end of the experiment showed some differences, both between lines and compared to the parameters used for index construction.
