The objective of this investigation was to select the proper model for genetic parameter estimation, to estimate genetic parameters, and to predict breeding values for litter size in most common sheep breeds in Czech Republic. A total of 143 896 lambing records from 1990 to 2012 were analysed. Variance components and genetic parameters for litter size were estimated separately for each breed using the BLUP animal model with repeatability. The basic model equation contained ewe age as a fixed effect and random effects of contemporary group, permanent environment and direct additive genetic effect of the animal. Modifications of the basic model were examined when various combinations of mating effects were included (contemporary group of ewes during mating (harem), additive genetic and permanent environmental effect of service ram). Estimates of phenotype variance increased across breeds (0.236 for Šumava to 0.779 for Romanov) with increasing breed average for number of lambs per litter. Variance component estimates for permanent environmental effect of the ewe were low (0.0001 to 0.0262). The variance of common environment of contemporary group (from 0.0223 to 0.1309) had bigger influence on the total variability of litter size in almost all studied breeds then additive genetic variance (from 0.0146 to 0.0587). The lowest heritability and repeatability estimates were for the Šumava (h2=0.0619; w2=0.0823) and Romney breeds (h2=0.0626; w2=0.0811); while the highest were for Merinolandschaf (h2=0.1091; w2 = 0.1129). The effect of service ram ranged from 0.01 to 0.02 of phenotypic variance in Šumava sheep, in Romanov sheep it was from 0.05 to 0.10, and in Suffolk sheep it was from 0.04 to 0.05. Including effects of mating (service sire, harem, and/or ram´s permanent environmental effect) in the model decreased deviance information criterion, what means that these models are more proper than the basic one. Results from present study demonstrate that genetic parameters did differ among the investigated breeds, which should be taken into account in breeding value estimation. The service rams have low but a clearly detectable influence on litter size of their mates. Genetic parameter estimates indicate that direct selection on the service ram effect could increase litter size and achieve genetic gain through ram selection.