As the popularity of robotic milking continues to increase, so has the need to breed cows that exhibit traits that are directly related to maximizing the return of the robotic milking unit. Selection based on milk output includes not only high production, but also fast parlor throughput. However, previous research has associated faster milking cows with increased levels of clinical mastitis. Therefore, the key is understanding the relationships between milk output components and production and udder health to produce the most parlor-efficient cows. Since 2009 AgSource (Verona, WI) has utilized Tru-Test (Tru-Test Inc., Mineral Wells, TX) electronic milk meters (EMM). EMMs are calibrated and used to collect monthly DHI milk weights, milking durations and milk samples. Using actual measures of milking speed (MS) from the EMMs removes bias introduced by the subjective visual classification of cows into MS categories. To investigate the impact of MS on areas of production and health, test day records (n = 681,029) were extracted from the AgSource DHI database. Data were used from cows with complete individual and sire IDs, and if at least 2 records existed where milk duration was less than 20 min, and MS less than 9kg/min. When categorizing cows into MS classes by increments of 0.5 kg/min, SCC was high for slow and fast milkers, but lowest for those milking 2.5 kg/min. Correlation between MS and ME305 milk was 0.9963, and between MS and average somatic cell count (SCC) was −0.0986. An association study was conducted to identify potential markers associated with MS. Genotypes on cows within the AgSource database were imputed to 50K using BEAGLE, with a final set of 52,890 genotypes for 1,326 Holstein cows available for analysis using the SNPassoc package in R. One marker was identified as significantly (FDR-corrected P < 0.03) associated with MS. The marker, located on BTA28 within gene KCNMA1, is a key regulator of smooth muscle contractions, and has been previously associated with breast cancer proliferation in humans. This candidate gene could be potentially beneficial in marker-assisted selection schemes to identify cows ideal for an automatic milking system.