Gross rainfall, net rainfall and stemflow were measured in a mixed deciduous woodland in southern England over a period of 14 months. Continuous measurements of standard weather data and momentum and sensible heat fluxes between the forest canopy and the atmosphere accompanied the investigation. The gross rainfall was corrected for catch losses due to high turbulence. Reliable net rainfall data were obtained from a combined application of simple storage gauges and troughs connected to automatic tipping bucket gauges. The evaporation rates from the wet canopy were calculated with the Penman-Monteith equation using the measured aerodynamic conductance to the momentum flux and, additionally, with the eddy covariance energy balance approach. Both methods agreed in the observation that the average wet canopy evaporation rate was slightly higher in the leafless period, due to higher wind speeds and the different aerodynamic properties of the canopy. Together with the lower average rainfall rate this counterbalanced the reduced storage capacity of the leafless canopy and maintained a relatively high interception loss throughout the year being 29% of the gross rainfall in the leafed period and 20% in the leafless period. The analytical sparse canopy rainfall interception model of Gash et al. [Gash, J.H.C., Lloyd, C.R., Lachaud, G., 1995. Estimating sparse forest rainfall interception with an analytical model. J. Hydrol. 170, 79-86] was parameterised, for the first time, for a mixed deciduous woodland. Separate parameters were derived for the leafed and the leafless canopy. The model explained the seasonal variability in the interception loss very well and is a suitable tool to analyse and predict this important component of the annual water balance of deciduous forests.