A method is presented for estimating the value of increased weighing precision when selecting pigs for delivery to the slaughterhouse. A dynamic programming approach is taken using hierarchic Markov processes in combination with a probabilistic growth simulation model. A heuristic delivery policy is optimized. Pigs with observed live weight higher than a threshold weight are delivered for slaughter. The pen is emptied depending on the age and number of remaining pigs in the pen. The optimal delivery policy is relatively robust to variations in weighing precision. The value of high weighing precision decreases markedly with increasing number of pigs per pen and is generally low. With 16 pigs per pen the low value of improved weighing precision leaves little financial room for investment in weighing equipment.