High-pressure fog (HPF) systems have advantages for greenhouse cooling compared to traditional systems, such as pad and fan. Such advantages include the potential of improving climate uniformity. Water is distributed throughout the greenhouse space thus reducing water use and energy operation costs, especially if used within naturally ventilated greenhouses. Fog cooling in combination with natural ventilation is difficult to manage, primarily because accurate estimation of air exchange rates is required to determine the precise amount of fog required. This limitation on automated control has been the main reason restricting the widespread commercial use of HPF systems. The goal of this research was to develop and implement a control strategy for a naturally ventilated greenhouse with a variable HPF system. The strategy that was developed included variable rate of fog introduced into the greenhouse, a dynamic control of the air ventilation openings, and it considered the contribution of cooling and humidification from the crop by evapotranspiration. Three evapotranspiration models, including Penman-Monteith, Stanghellini and Takakura, were calibrated and evaluated in terms of prediction accuracy. The Stanghellini model provided the best overall performance for several growing seasons and under two different evaporative cooling systems (i.e. pad and fan and natural ventilation with HPF), and was selected and implemented in the cooling control strategy. The strategy utilized enthalpy and vapor pressure deficit (VPD) of the greenhouse atmosphere for the control parameters. Using a calibrated greenhouse mechanistic climate model, a computer algorithm was created to simulate the capabilities of the proposed. The control strategy that was developed was able to maintain the greenhouse climate closer to the pre-established set points while consuming less water and energy, compared to a constant HPF system based on VPD control. Finally, the strategy was implemented in a single span research greenhouse. A four-day validation study provided good agreement for measured and simulated greenhouse climate values, as well as for water and energy use. Moreover, the strategy was able to maintain VPD around its set point for all the experiments and temperature remained around its set point when outside enthalpy was lower than the enthalpy set point.