Dietary intake data of 43 Finnish rheumatoid arthritis patients were collected using 7-day food records. The subjects were randomized into a control and a vegan diet groups, consisting of 22 and 21 subjects, respectively. The subjects in the vegan diet group received an uncooked vegan diet ('living food') for 3 months, and they were tutored daily by a living-food expert. The subjects in the control group continued their usual diets and received no tutoring. Adherence to the strict vegan diet was assessed on the basis of urinary sodium excretion and by the information on consumption of specific food items (wheatgrass juice and the rejuvelac drink). The use of these drinks was variable, and some boiled vegetables were consumed occasionally. However, only one of the subjects in the vegan diet group lacked a clear decrease in urinary sodium excretion. Rheumatoid patients had lower than recommended intakes of iron, zinc and niacin, and their energy intake was low compared to mean daily energy intake of the healthy Finnish females of the same age. Shifting to the uncooked vegan diet significantly increased the intakes of energy and many nutrients. In spite of the increased energy intake, the group on the vegan diet lost 9% of their body weight during the intervention period, indicating a low availability of energy from the vegan diet.