The periodic need to restock reagent pools for genotyping chips provides an opportunity to increase the number of single nucleotide polymorphisms (SNP) on a chip. As an improved replacement for the GeneSeek Genomic Profiler HD for dairy cattle, a set of >140,000 SNP was selected that included all SNP on the current chip, all SNP used in genomic evaluations, SNP that are possible functional mutations, and other informative SNP. Most added SNP were selected from the Illumina Bovine HD Genotyping BeadChip based on the magnitude of effects on evaluated traits. Some SNP with lower minor allele frequency were considered because of their potential for better tracking of causative variants. Genotypes already available from other chips were used to impute and evaluate the SNP set. Effects for 134,511 usable SNP were estimated for all breed-trait combinations; SNP with the largest absolute values for effects were selected (5,000 for Holsteins, 1,000 for Jerseys, and 500 each for Brown Swiss and Ayrshires for each trait), which resulted in 78,032 SNP after removing duplicates. An additional 9,130 SNP with many parent-progeny conflicts after imputation were removed, which resulted in 72,843 SNP. Of those, 38,515 were among the 60,671 SNP currently used in genomic evaluation. To minimize possible accuracy loss,12,094 of the SNP currently used but not already selected and with the largest effects were added for a total of 84,937 SNP. Three cutoff studies were conducted with 60,671, 84,937, and 134,511 SNP to determine gain in reliability over parent average when evaluations based on data from August 2011 were used to predict genetic merit from December 2014. Across all traits, mean gains were 32.5, 33.4, and 32.0 percentage points, respectively. Previous experience indicates that gains from the highest number of SNP will increase as the number of genotypes from the new SNP set increases. The gain of 0.9 percentage points from adding nearly 25,000 SNP justifies the extra computation time needed. However, the gain may be overestimated because data used to select the most informative SNP were also the data used to determine gain.