A mechanical dewatering process was developed to avoid losses in yield and quality incurred during field hay curing. A 4‐yr field study was conducted to gain information on factors affecting hay curing losses and to compare hay to mechanically dewatered alfalfa (Medicago sativa L.). Ten harvest regimes were imposed on alfalfa monocultures to provide a wide range of fresh forage compositions and of hay curing conditions. Over the 4 yr, hay averaged 31.0 g kg−1 of N, 3.3 g kg−1 less than herbage, but 3.7 g kg−1 higher than pressed forage. Hay in vitro dry matter disappearance (IVDMD) was 596 g kg−1, 10% less than herbage, but similar to pressed forage at 589 g kg−1. Pressed forage was highest in neutral detergent fiber (NDF) at 559 g kg−1 while fresh forage had 422 g kg−1 and hay had 518 g kg−1. The best single predictors of hay concentrations of N, NDF, and IVDMD were fresh forage concentrations of the same constituents. Other significant variables for prediction of hay composition included dry matter loss (DML) during hay curing, rain amount, and hay curing time. Since hay curing and pressing both affected NDF concentration more than IVDMD, preservation system effects on intake potential could be even greater than effects on digestibility. The finding that fresh forage quality is the primary determinant of alfalfa hay quality indicates that close attention to quality at the time of cutting should more consistently produce hay of the desired quality than would delaying harvest in an attempt to avoid rain. The pressed forage and the protein concentrate together produce yields similar to or greater than those of hay and reduce N losses by nearly one‐half.