Twenty-four lactating Holstein dairy cows (12 first lactation and 12 multiparous; day in milk = 11 ± 5 days) were allotted to a randomised complete block design in a 2 × 3 factorial with four replicates per treatment to evaluate the effects of two methods of alfalfa feeding (dry and reconstituted to achieve a theoretical dry matter (DM) content of 350 g/kg) and three geometric mean (GM) particle sizes of alfalfa (9.13, 4.51 and 1.20 mm) on performance of dairy cows for a period of 28 days. Diets were offered for ad libitum intake as total mixed rations (TMR). The GM particle size, its standard deviation, and the values of physical effectiveness factor of alfalfa and TMR decreased as alfalfa particle size decreased. Reduction of particle size and reconstitution of alfalfa increased the bulk density and the functional specific gravity of alfalfa and rations. Reduction of particle size decreased insoluble dry matter, water-holding capacity, and hydration rate of alfalfa. As particle size decreased, the amount of physically effective NDF in the ration (g/kg) decreased but the daily intake of physically effective NDF (kg/day) increased. Reduction of particle size and reconstitution increased dry matter intake (DMI) and ruminal passage rate, but reduced NDF and ash digestibilities, ruminal pH, N-NH3, milk fat, total chewing activity, rumination and eating time, total and ruminal mean retention time, and time delay of marker. Increased functional specific gravity, from reduced forage particle size and the reconstitution of alfalfa, was the most important factor influencing DMI, milk composition, and chewing activity.