This paper presents a simulation study that shows that failure to fit dam permanent environmental effects in variance component estimation of lamb survival results in an upward bias in the estimate of maternal genetic variance. In contrast, fitting litter effects has little impact on variance component estimation. These results have implications for sheep genetic evaluation of lamb survival, and values for the direct and maternal heritability of lamb survival at the lower end of the range of those published based on alternative models should be used in national genetic evaluations.