Dry period lengths (DPL) are known to be around 60 d but may systematically vary between and within herds. These systematic variations are not well known. The objective of this study was to describe the distribution of DPL among Holstein herds across the United States and to investigate the association between parity, days to conception, and seasons with DPL. After editing, DHIA lactation records from cows in their first and second parity (n = 438,087) were available for 1,167 herds. Cows calved from January 2012 to 2016. To be included in the analysis, herd size was at least 50 cows, cows completed their first lactation and had a subsequent calving date, and length from conception to drying off was at least 180 d. Daily milk yield was individually quantified by fitting Wood’s lactation curve to milk test records. The variable of interest was days from conception to dry off (CtoD) because this measure was less affected by irregular gestations lengths, and hence a better indication of the intended DPL. Four seasons were defined based on date of dry off. The association between season and CtoD was calculated as the largest difference in CtoD among the 4 seasons. Factors explaining within herd variations in CtoD were investigated with simple linear regressions of CtoD with days to conception and parity, by herd. Overall, cows in the first parity had longer CtoD but less variation (224.4 ± 14.31 d) than cows in the second parity (220.3 ± 17.10 d). Compared with second parity CtoD, 2.5% of farms had ≥ 5-d shorter CtoD for first parity cows, whereas 43% of herds had ≥ 5-d longer CtoD for first parity cows. In 15% of the herds, the mean difference between CtoD in first and second parity cows was at least 1 d. In 34.1% of herds, the regression of days to conception on CtoD was significant (P < 0.01). In 4.3% of the herds, each later 10 d of conception increased CtoD by at least 1 d. In 18.5% of herds, the regression of season on CtoD was significant (P < 0.01). In conclusion, CtoD was typically greater for first parity cows than second parity cows, indicating likely shorter DPL in first parity cows. Days to conception and season all were associated with CtoD in a large number of farms, which indicates that many dairy farmers vary cows’ DPL on purpose.