Circadian variations in plasma iron levels were first reported in humans in 1937. Influences of the sleeping pattern and of plasma cortisol and adrenaline levels on these variations as well as the reproducibility of the phenomenon itself are discussed controversially in the literature. The influence of food intake, however, was not considered in most of the studies and is therefore subject of this investigation. Circadian plasma iron and plasma transferrin variations were determined in rabbits and compared under free access to food and under starvation (caecotrophy was not prevented). Population-mean-cosinor analysis of circadian plasma iron concentrations showed similar variations in the fed and starved condition (mesor: 116.6 micrograms/dl vs 118.1 micrograms/dl, acrophase 0752 hr vs 0728) except for a significant increase of the circadian amplitude under free access to food (30.9 micrograms/dl vs 22.3 micrograms/dl, P less than 0.05). There was no variation in plasma transferrin, which shows that 24 hr variations in plasma iron are not caused by modulation of plasma transferrin. These findings demonstrate a circadian rhythm for plasma iron, the amplitude of which is increased by food intake.