As long as the World Health Organization continues to report that more than 250,000 cases of paralytic poliomyelitis occur annually in developing countries, the United States must be prepared for periodic exposures to wild virus.' Until poliomyelitis is brought under control throughout the world, it behooves us to maintain a high level of vaccine-induced immunity. Hinman and his colleagues from the federal Centers for Disease Control, in the first paper in this series,2 document how well this has been achieved in the nation through the widespread use of OPV. From more than 20,000 cases of paralytic poliomyelitis in 1952, we now have a situation where endogenous wild poliovirus seems to have disappeared from the United States; and when the few imported cases are considered, there is virtually no spread of virus from them. Unfortunately, OPV, like other vaccines, is not perfect, although it is more nearly perfect than most. About five or six vaccine-associated cases now occur per year. Over the past 15 years more cases have occurred in contacts, particularly household contacts, than in the vaccinees. This is not surprising since it was known even before licensing that the virus excreted by vaccinated children was less attenuated than the vaccine itself, having regained some neurovirulence in monkey tests.3 At that time, some recommended that families be vaccinated as units, to avoid exposing susceptibles to mutated virus from the vaccinee. Had that been done, contact cases would have been avoided. The issues raised by the foregoing Hinman/Salk papers2'4 hinge on the possible shift in immunization policy from OPV to IPV. One of the public health axioms on which I was brought up is that if a large-scale program is working wellsuch as reducing the burden of paralytic polio from 20,000 cases per year to less than 10 per year-it should not be tinkered with unless we are certain that the resultant effect will be an improvement. Ifwe had used IPV, it is doubtful that we would have been as successful as the Dutch and the Scandinavians-and even those homogenous populations with their excellent vaccine coverage have had some outbreaks during recent years which, if extrapolated to our large population, would have meant many hundreds of paralytic cases and over a million carriers of wild virus. There is reason to consider a properly timed combination of both IPV and OPV. Such a program would not have any possibility of raising the risk of OPV-associated poliomyelitis and indeed may eliminate cases altogether. Combined IPV/OPV vaccine regimens have proven successful in the tropics where the local population had been subjected to regular and high importation of virus from neighboring countries, resulting in frequent wild virus challenge very early in infancy5'6 In such environments, the killed poliovac-