A national genetic evaluation program for hoof health could be achievable by using hoof lesions collected directly by hoof trimmers. However, not all of the cows present in the herds during the trimming period are usually presented to the hoof trimmer and the pre-selection is rarely random, leading to potential bias and, consequently, inaccuracies in the genetic evaluation. The objective of this study was to investigate a multiple trait evaluation to account for the potential bias due to pre-selection of cows for hoof trimming. Hoof lesions from 70,305 animals were recorded by 23 hoof trimmers in 521 Canadian herds from 2009 to 2012. The lesions included in the analysis were digital dermatitis, interdigital hyperplasia, and sole ulcer. Multiple trait evaluation was performed to account for the cow pre-selection bias. In addition of the presence of the lesion, locomotion (LOC) and the overall score for feet and legs (FL) were considered. The differences between average EBV of the lesions estimated with univariate or multiple trait model increased as the percentage of non-trimmed daughters increased, suggesting that including LOC and FL might have an effect on the estimations. However, correlations between the EBV estimated with single or multiple trait models by percentage of not-trimmed daughters were very high (0.97 to 0.99), showing that the sire ranking was very similar.