The in#uence of the greenhouse ventilation regime (natural or forced ventilation) on the energy partitioning of a well-watered rose canopy (Rosa hybrida, cv. First Red) was investigated during several summer days in warm mediterranean conditions (Volos, Eastern Greece). Two types of ventilation systems were evaluated: (i) fans and roof openings (forced ventilation) and (ii) roof openings only (natural ventilation). The relevant climatic variables (air temperature and vapour pressure de"cit), average canopy temperature, transpiration rate and intercepted net radiation were continuously monitored. The data were processed in order to derive the canopy-to-air sensible heat #ux, the bulk aerodynamic and stomatal conductance of the canopy, and the Bowen ratio. The results indicated that forced ventilation induced a more homogeneous vertical "eld for temperature and humidity, by providing a more intensive mixing of the inside atmosphere. It was found that the canopy-to-air sensible heat #ux and the transpiration rate were not signi"cantly enhanced under fan-assisted ventilation and that the Bowen ratio was similar under both regimes, ranging from !0)5 to !1)0. The canopy-to-air temperature di!erence was signi"cantly di!erent, being less negative under forced ventilation. Calculated values of the crop water stress index were similar and relatively high (+0)5) for both regimes. The values of the bulk aerodynamic conductance were found to be dependent on the ventilation regime, those for forced ventilation being roughly double than those for natural ventilation. The estimated values of the stomatal conductance of the canopy were slightly higher (about 10}15%) under natural ventilation. Normalizing the stomatal conductance of the canopy by radiation revealed a strong stomatal response to the humidity conditions prevailing at the surface of the canopy. It appeared that forced ventilation increased signi"cantly the aerodynamic conductance, but did not in#uence signi"cantly water consumption when compared to natural ventilation, because of the negative feedback between canopy-to-air vapour pressure de"cit and stomatal conductance.