Evaluations of the economically optimal dry period length (DPL) are scarce. Our hypothesis was that optimal DPL depends milk and feed prices as well as on the most limiting herd constraint such as a the number of milking cows (MLK), total cows (TOT), milk quota (QUO), or total milking time (TIM). The objective of this study was to quantify the optimal DPL under these 4 herd constraints, as well as provide a sensitivity analysis based on milk and feed prices. A model with 2 adjacent lactations with 31 weekly categories of DPL (from 0 d to 210 d) between the 2 lactations was build. Standard Wood’s lactation curves were used. Milk loss and days open in the subsequent lactation as a result from varying DPL were predicted by polynomial regression equations fitted on data from the literature. Income over feed cost (IOFC) per herd per day was the measure of interest. In the default case, all constraints were met simultaneously. Milk price was $0.35/kg and feed costs were $0.26/kg DMI for lactating and $0.08/kg DMI for dry cows, respectively. Default days open was 120 in each parity. Culling was not included. These inputs allowed for the calculation of the number of lactating and dry cows in both parities from which IOFC were calculated. Default results showed that the optimal DPL under 4 constraints were 98, 77, 98, and 91 d for MLK, TOT, TIM and QUO, respectively. MLK and TIM constraints resulted in 19% of all cows dry whereas TOT had 16% of cows dry. A 56 d DPL let to losses of $0.40/milking cow per d (MLK), $0.09/cow per d (TOT), $0.56/hour of milking time per d (TIM), and $0.29/100 kg milk (QUO) compared with the optimum DPL. Greater milk price and lower dry cow feed costs increased the optimal DPL. Adding other variable costs reduced the optimal DPL for MLK. In conclusion, the optimal DPL depended on herd constraint as well as milk and feed prices. Longer DPL for otherwise low producing cows allows for milking more higher producing cows when MLK and TIM were herd constraints. Cow performance responses to varying DPL as well as inclusion of other costs may need to be refined for final results.