In 1958 and 1959, 7,333 residents of 28 villages and 2 boarding schools in the Bethel area of Alaska were enrolled in a controlled trial of the effectiveness of isoniazid in preventing tuberculosis. Of those who were eligible for medication, 3,017 took placebo only and 3,047 isoniazid only. The latter was prescribed in a dosage of roughly 5 mg. per kg. of body weight, given once daily for a year. Approximately 70% of the recommended annual dosage was taken by participants in both groups. Complaints attributable to the side effects of isoniazid caused 0.8% of the population taking that drug to discontinue medication. Gastrointestinal symptoms and dizziness were the only complaints more common among persons taking isoniazid than among those taking placebo. Large overdoses of isoniazid are known to have been taken by 5 children. Two died; the other 3 recovered completely. Experience in the trial indicates that accidental poisoning by isoniazid is preventable. During the year of medication, isoniazid prevented approximately 70% of the serious manifestations of tuberculosis in the participating population. The period of observation following the medication year is still insufficient for definite conclusions, but the tuberculosis rates among persons who took isoniazid are lower than among these who took placebo. If no one in this study population had taken prophylactic isoniazid, 173 serious manifestations of tuberculosis would have been expected during the period since the initiation of the trial, contrasted with an estimated 88 if all participants had taken isoniazid. It appears that prevention is less costly than cure.