Previously, we have shown that cohort composition had minimal effect on the computation of composite selection tests when the breeds with small sample size (n < 10 and n < 20) were excluded (Randhawa et al. 2014). Further, in this study, we also assessed the effect of cohort composition comprising European breeds of cattle, specifically whether a few breeds with large sample size (n > 20) can disproportionately affect the analyses. To this effect we performed additional analyses by limiting the sample size (n ≤ 10 and n ≤ 20) per breed. These results were compared to those from the original cohorts, where we included all available samples from each breed and the constituent and composite selection tests were computed by analyzing the European large and small cohorts against each other.