The wind shadow model has been developed to calculate the wind sheltering effects of upwind obstacles for air infiltration calculations. This effect must be determined for infiltration calculations because, in almost all situations, only the unobstructed mean wind speed is known for a building site. This model has adapted the theoretical calculation procedures developed for far wake centreline velocity deficit calculations to near field flows, where shelter has a significant effect. The model uses the concept of a wind shadow projected downstream by upwind buildings to determine the effect of wake velocity reduction on building surfaces. The turbulent nature of the wake is accounted for by "flapping" the wake over a range of wind directions. The effectiveness of this model in accounting for sheltering effects in infiltration calculations has been examined by comparing infiltration model predictions including the wind shadow model to measured data from a row of test houses. The measured data covered a wide range of wind speeds, wind directions and leakage distributions by using over five thousand hours of infiltration measurements from five houses.