SUMMARYA trial has been conducted of the following methods for the reduction of serum cholesterol level in patients with atherosclerosis: (i) The use of a low‐fat, low‐cholesterol diet with the addition of corn oil; (ii) the exhibition of triparanol (0·25 to 0·5 gramme per day); (iii) the administration of nicotinic acid (up to 3 grammes per day); (iv) the administration of a combination of nicotinic acid and triparanol. The results were as follows.With the dietetic method, a fall in plasma cholesterol level usually occurred. However, this was only of very modest degree (mean of 42 mg. per 100 ml. in 10 patients). Although a proportion of the patients treated by diet considered that their vascular symptoms were relieved, this could not be confirmed by step tests or by clinical examination.When triparanol was used, significant lowering of the plasma cholesterol level (determined by the Bloor method) was observed in 17 of 22 patients treated for six months (mean lowering, 69 mg. per 100 ml.). Desmosterol accumulated in the plasma, and the true lowering of the cholesterol level was somewhat greater than this; the lowering of the total plasma sterol levels was somewhat less. We were unable to demonstrate desmosterol in the intimas of patients treated with triparanol. Treatment was not associated with any clinical benefit, nor were there any side effects which could be attributed to the drug.The administration of nicotinic acid was associated with a moderate lowering of the plasma cholesterol level (mean of 38 mg. per 100 ml. in 10 patients). Although a few patients said that they had experienced clinical benefit, this was of doubtful significance. Side effects were common and prevented the administration of a full dose in most cases. No toxicity was demonstrated.The use of a combination of nicotinic acid and triparanol produced an additional fall in the plasma cholesterol level (mean of 60 mg. per 100 ml. in nine patients). The addition of triparanol treatment to nicotinic acid treatment also produced an additional fall (mean of 44 mg. per 100 ml. in five patients).These findings are discussed, and it is concluded that it is possible to lower the plasma cholesterol level by diet or drug treatment, but that this is not of great clinical value if it is employed for a relatively short period or in the treatment of elderly patients. Treatment is more likely to be of value in younger men with early atherosclerosis or a raised plasma cholesterol level, but, as treatment must be carried on over many years, the method used must be devoid of serious toxicity.