Boston, Massachusetts Desmond Laurence introduced the notion that a therapeutic agent or approach, like the process being treated, has a natural history [l]. Information on the efficacy of an agent and any adverse reactions-in a relatively limited number of patients-are compiled for submission to a regulatory agency. When efficacy is established at a cost in adverse reactions that is sufficiently low, the agent is marketed. Few would doubt that the rigorous demands of the regulatory agencies have provided an important element of protection for the community in this country and elsewhere. On the other hand, it is clear that the documentation required for the regulatory agency cannot provide all the information physicians and patients require. For that reason, presumably, there are several recent examples of effective agents withdrawn shortly after their release because of’unanticipated and initially unrecognized adverse reactions. The studies, designed to satisfy the regulatory agency, may not be sufficiently large to identify low frequency events; are often designed to preclude potential drug interactions, so these too will be missed; and they may exclude patient groups that have a specific, unanticipated sensitivity to the potential adverse effects of a therapeutic agent. For that reason, postmarketing surveillance should become an increasingly important part of a therapeutic agent’s natural history. Maxzide, a combination of triamterene and hydrochlorothiazide, was first marketed about a year ago, having received approval from the Food and Drug Administration because of two features: first, the combination was formulated to make bioavailability optimal; and second, the ratio of triamterene to hydrochlorothiazide was documented to reverse thiazideinduced hypokalemia [2,3]. The PACT study was designed primarily to provide information on the adverse reaction rate via a prospective, systematic post-marketing surveillance program. The study was initiated in March of 1985 with an ultimate goal of enrolling 75,000 patients by December 31, 1985. This report, on the first 47,465 patients entered, was prepared for the symposium that accompanies it and must be viewed as an interim analysis, written primarily because of insights gained that are potentially important to the medical community. Even the interim report, reflecting experience in more than 47,000 patients, is the largest such study ever reported in the United States.