The aim of this study was to evaluate the effect of shortening or omitting the dry period of dairy cows on milk casein composition. For this study, we analyzed milk samples of 90 cows with a dry period of 0, 30, or 60 d and either a glucogenic or a lipogenic ration in early lactation. Milk was sampled at 6 and 2 wk prepartum and at 2, 6, and 12 wk postpartum. Milk was analyzed for casein (CN) composition by capillary zone electrophoresis, and isoforms of κ-CN were measured by reversed phase-HPLC. Shortening the dry period from 60 to 30 d reduced the αS1-CN fraction by 3.8% and increased the αS2-CN fraction by 5.5%. In milk from cows with a 0-d dry period, the glycosylated κ-CN fraction in late lactation increased from 8 to 12% between 6 and 2 wk prepartum. After calving, the glycosylated κ-CN fraction in milk was higher for cows with a 0-d dry period (6.7%) compared with cows with a 60-d dry period (5.2%). The glycosylated κ-CN fraction at 2 wk postpartum was negatively correlated with milk yield, suggesting that glycosylation was related to reduced productivity of mammary epithelial cells. In early lactation, the β-CN fraction was reduced in milk of cows with a 0-d dry period. A lowered β-CN fraction was associated with high somatic cell count and greater parity, indicating that it was the result of proteolytic activity. In conclusion, casein composition changes that result from shortening the dry period from 60 to 30 d are not expected to affect processing characteristics of milk. Applying a 0-d dry period may affect processability of milk because of a higher glycosylated κ-CN fraction, and possibly because of higher proteolytic activity compared with a 60-d dry period.