Abstract This article presents numerical simulations of the distribution of climate parameters within a ventilated tunnel tomato greenhouse during variable outside conditions. The simulations were performed using a computational fluid dynamics (CFD) code that solved the transport equations in a 3D domain, including the greenhouse and its crop stands, the surrounding ambient air and the soil located directly under the greenhouse. Radiative heat transfers were modeled using a bi-band discrete ordinates (DO) model, and the crop was considered to be a porous medium. Sensible and latent heat transfer between leaves and the surrounding air were determined based on the energy balance that included longwave and shortwave radiation fluxes in each crop control volume. The climatic boundary conditions were determined using experimental measurements, and the sun position was calculated for each time interval that was considered. The temperature distribution in the soil was determined based on a preliminary CFD determination of the conductive heat transfer in a 1D soil column. Simulations in the entire 3D domain were then performed, and 1 h time step and boundary conditions were updated prior to each calculation procedure. Results are presented for the spring equinox and summer solstice. These results highlight the combined influence of sun position, wind direction and intensity on the greenhouse microclimate and especially on the evapotranspiration rate of the crop at the leaf level. We discuss the possibility of using CFD code integration as a conceptual tool for designers or in association with a control climate model for farmers.