The efficiency of two simplified methods of selecting litter size in rabbits has been studied. These methods, a family index and a reduced BLUP using only records of two generations, have been compared with a BLUP, repeatability-animal model that used all records from the foundation until the generation currently selected. The comparison has been made analyzing correlations between rankings of matings, ratios of coincidence of matings selected to contribute progeny to the next generation, and ratios of response lost when the simplified methods are used. An analysis of robustness of the methods to changes in genetic parameters was also carried out. Records of 18 and 14 generations of two lines of rabbits selected for litter size at weaning were used, and the selection process was studied in generation 17, 18 of one line, and 13, 14 of the other. The two simplified methods were practically similar in the context of our study, with non overlapping generations and good balance between animals to be selected and fixed effects. The correlations between rankings of a simplified method and the BLUP using all records ranged between 0.79 and 0.82, the coincidence ratio between 0.62 and 0.90, and the loss ratio between 0 and 0.17, being the average around 0.05-0.06. The methods of selection revealed themselves as very robust.