Relamping growth chambers is key to provide dimmable and a quality light source to plants, in addition to reducing electric energy consumption. Here, relamping was performed in PHENOPSIS, a high throughput phenotyping platform dedicated to the study of the genetic variability of plant response to contrasted environmental conditions, including soil water deficit, high temperature, light and CO2. Each of the three growth chambers of this automated platform allows the cultivation of up to 504 plants in highly controlled environmental conditions. These 6 m2-growth chambers were originally equipped with 400 W mercury iodide (hydrargyrum quartz iodide, i.e., HQI) lamps and fluorescent tubes. The main objective was to provide a dimmable light with homogeneous intensity and quality to plants in order to reduce phenotypic variability within the growth chambers and increase repeatability among experiments. Sunlike LEDs were used in preliminary experiments performed in smaller growth chambers in order to test the stability of the phenotypes of Arabidopsis thaliana and durum wheat (Triticum turgidum subsp. durum) (dry weight, leaf numbers, leaf area, stomatal density, flowering date, IR-determined leaf and soil temperature). After installation in PHENOPSIS, light physical characteristics (PAR, spectrum) and plant morpho-physiological traits under Sunlike LEDs were mapped. The three watering levels showed high plant phenotypic stability of A. thaliana within the growth chamber. We report a significant reduction of electric consumption in the new LED-equipped lighting system compared to the previous one, all other parameters being similar. This equipment will allow testing the role of light intensity in plant response to various environmental stresses.