Experiments on plants are often carried out in growth chambers or greenhouses which necessitate the use of an artificial rooting environment, though this is seldom characterized in detail. Measurements were made to compare the rooting environment in large boxes (0.25 m3) with that in small pots (0.19, 0.55 and 1.90 dm3) in naturally lit chambers. Diurnal temperature fluctuations of 14.6, 11.6 and 7.7°C occurred in the post compared with only 1.9°C in the boxes. Soil drying to a matric potential of-50 kPa was approximately 25 times faster in the pots. The mean heights of 2 year old Sitka spruce (Picea sitchensis (Bong.) Carr.) seedlings grown throughout their second growing season in the three sizes of pots were 38, 62 and 92% of the mean height of those grown in the boxes. Soil solution nutrient concentrations in the boxes were considerably increased by soil drying, an aspect which seems to have received little attention in experiments involving artificially imposed drought. An alternative system of constraining the roots of individual plants within nylon fabric bags, embedded in larger volumes of soil, to facilitate harvesting of complete root systems is described. The importance of the rooting environment in determining the outcome of physiological experiments is also briefly discussed.