The most widely used heating system in Israeli greenhouses is based on hot air, distributed in the greenhouse via perforated polyethylene ducts. Recently, there has been a growing trend to install hot-water pipe systems in new greenhouses. In both systems, an on–off controller is generally used to control the air temperature within the greenhouse. Experiments were conducted to compare the effects of the two heating methods on greenhouse air temperature, humidity and crop temperature, and on the energy required to keep the air temperature in the greenhouse at a desired level. The experiments were done only during the night, in experimental and commercial greenhouses in which roses were grown. The results showed that, with both heating methods under on–off control, there are cyclic changes in the temperatures of the crop and air, the temperature difference between crop and air, and humidity ratio of the greenhouse air. Fewer heating cycles per night were observed with pipe heating than with air heating, because of the relatively large thermal inertia of the water-filled heating pipes. The humidity ratio at crop level increased during the heating up period of each heating cycle in both heating methods, but the rate of increase and the amplitude were larger with air heating than with pipe heating. The radiation heat transfer between the heating pipes and the crop ensured that, with pipe heating, the leaves in the lower part of the crop (where the pipes were placed), especially those facing the pipes, were generally warmer than the air, and those in the upper part colder than the air, during heat up. With air heating, the leaves were cooler than the air in both the upper and the lower regions of the crop during most of the heating cycle. The results show that there is no significant difference between the two methods in the energy consumption required to obtain a given temperature level inside the greenhouse, as long as the pipes and ducts are positioned between the plant rows.