Survival of commercial laying hens is an important trait. Feather pecking has a large effect on the survival of birds. To improve survival it is important to use quantitative genetic methods that take into account both the direct genetic effect (victim effect) and the indirect genetic effect (actor effect). For survival time, we found that the victim effect contributes 13-64% of total heritable variation, while the actor effect contributes 36-87% of total heritable variation. Together, they explain 15-26% of total phenotypic variation in survival time. Here we compare different breeding programme designs to identify the optimum selection strategy against mortality due to feather pecking. Results show that mortality due to feather pecking can be reduced using genetic approaches, taking into account direct and indirect genetic effects. We performed a selection experiment using selection based on relatives. Using this method enables selection against mortality due to feather pecking in laying hens. However, selection intensities were small. Genomic selection can be a promising tool to select against mortality due to feather pecking. Model predictions show that genomic selection is expected to yield a rapid reduction of mortality due to feather pecking, but a challenge will be to reduce mortality due to feather pecking in large groups.