AMIPHENAZOLE (2 : 4-diamino-5-phenylthiazole hydrochloride) was shown by Shaw and Bentley (1952) to be one of several substances that antagonized the effect of morphine in depressing respiration. Since then there have been several reports of the efficacy of parenteral amiphenazole for this purpose (Fleming, 1958; Webber, 1958) and also of its value as a respiratory stimulant in cases of acute respiratory failure (Simpson and Hawkins, 1959; Litde, 1962 ). The last author also tried to assess objectively the effect of the drug by measuring its effect on minute ventilation and pCO~. He and Simpson and Hawkins employed doses of 15o rag. of amiphenazole in most instances. To date there have been no reports assessing the effacy of the oral preparation supported by lung function studies. Indeed any such trials might have been complicated by a change in the recommended dosage. Amiphenazole is marketed in this country by Nicholas Laboratories under the trade name Daptazole and tablets of 2o mg. were first issued ill 1956 with a recommended dosage of 6o-I2O mg. a day. Further studies suggested a much higher dosage (comparable to that given parenterally) was needed to stimulate respiration and a stronger tablet containing IOO mg. (Daptazole ioo) was brought out in September I961. A r6gime of 300-800 rag. a day in divided doses is proposed and at this level the drug is claimed to increase ventilation and reduce the pCO~ in some patients with chronic respiratory failure. Increased ventilation is brought about by an increase in tidal volume rather than increase in respiratory rate. This study is an attempt to evaluate the value of oral Daptazole ioo in chronic respiratory failure. It is difficult to perform a controlled trial of any drug in padents with chronic respiratory failure for several reasons: first, they are usually receiving other forms of therapy such as breathing exercises, expectolants, antibiotics, antispasmodics, digitalis and diuretics; and secondly, the degree of infection and airways obstruction may vary from day to day. Such variables were minimized by initially treating patients with a variety of the measures mentioned above until the maximum response had occurred as judged by the clinical state and conventional simple pulmonary function tests (lung volumes, ventilation and blood gases). Once patients had reached this stage they had a run-in of several days a t this steady state before treatment was started. It has been possible to study only a limited number of patients (ten) under these controlled conditions. Most of them had a period of a week or more of study whilst taking