A mathematical model was developed to describe the rates of drying of perennial ryegrass across a number of di!erent treatments at mowing and after mowing and di!erent weights of grass per unit area. In the experiment, weight change of grass in wire-mesh trays was measured in the "eld over 2 d (32 h). Weight change was used to assess water loss and hence drying rate of the grass. The e!ects on drying rate of "ve weights of grass per unit area (1)5, 3, 6, 12 and 24 kg [fresh material] m), three treatments at mowing (no treatment, mowerconditioned, #ail-treated) and three treatments after mowing (no treatment, inverted, mixed) were examined in a 5;3;3 factorial experimental design. The experiment was replicated twice on each of 16 occasions in 1992 at the Agricultural Research Institute of Northern Ireland. This gave a total of 32 replicates per treatment. The trays were weighed at 2-h intervals from 09)00 to 17)00 h each day. Three potential evaporation parameters (net/solar radiation, the Penman equation) were compared for use in a "nal empirical model. Drying rate parameters were calculated for each of the individual treatments. Relationships were developed between drying parameters and weight of grass per unit area for each of the treatments at and after mowing. Inclusion of a rainfall correction factor made the prediction of grass dry matter concentration more complex. The "nal empirical model reliably predicted dry matter concentration of grass in trays in the "eld over 32 h across the wide range of weather conditions and mechanical treatments that were studied. This "nal model which incorporated a single evaporation parameter (solar radiation) was shown to describe the pattern of grass drying in the "eld as accurately as a "nal model based on the Penman equation. However, further work is required to establish the extent of any di!erences between grass drying rates in the wire-mesh trays used in the current work and grass drying rates from commercial-type swaths.