In 2014, the Netherlands and Flanders implemented a new genomic system, in which direct genomic values (DGV) are used as pseudo-observations to directly affect the conventional trait. For most trait groups, this pseudo-record system is a multi-trait animal model, where the added R² from a validation study is used as correlation between DGV and conventional trait, and a covariance matrix holding partial correlations is derived from it. However, validation of the DGV is done in a single trait setting, where a relatively young –and unreliable- group of bulls is used as validation set, often resulting in unstable added R², depending on which set of bulls you select as validation set. When the added R² of traits within a multi-trait system are not consistent to each other, the multi-trait system is unbalanced, even if the defined covariance matrix is positive definite, resulting in incorrect reliabilities. In this study, a consistency check is done on the validation results of a set of fertility traits (nonreturn56, interval-calving-1st-insemination, calving interval, interval-first-last-insemination, conception rate, conception rate heifers, and age at first insemination) used in a multi-trait animal model, to assess the correctness of the added R², the used correlations and the reliability calculated for the GEBV. This consistency check is based on the coefficient of multiple correlations (MCC); where the expectation is that the diagonal of the matrix containing MCC is equal to the corresponding R² from the validation. Data of the EuroGenomics consortium was used. The training set varied between 6,000 bulls (for traits with national data only) to 26,000 bulls. Bulls born after 20040630 were considered as validation set. The added R² ranged from 0.21 for age-at-first-insemination to 0.64 for interval-first-last-insemination. When the set of added R² was consistent, the reliabilities for the validation bulls estimated in the multi-trait model were similar to realized reliabilities. In conclusion, the consistency check is a valuable tool to assess genomic validation.