In that they were made in temperate countries, most of the studies on poliomyelitis may not apply in sub-continent of endemic infantile paralysis. This review brings together data on polio in India, to present any changes which may have occurred since 1940. Only about 2% of children with polio die in the acute illness; about 95% of all cases have paralysis of one or both legs. In lameness surveys the adjustments for deaths and for arm paralysis are unnecessary and inflate the prevalence. Surveys suggest that prevalence has risen, but there may be other explanations for the figures. Around sentinel centres with effective cold chains, prevalence has decreased rapidly. The figures of the National Baseline Prevalence survey are examined. There are few cases of provocation, but a new phenomenon of aggravation by unnecessary intramuscular injections given to children with fever has been described. Such unnecessary injections are thought to be the cause of more severe paralysis in about 45% of cases and of converting a non-paralytic attack into paralysis in another 30% of the perhaps 200,000 cases in India each year. Aggravation is thought to be caused by a mechanism similar to the effects of physical activity. It is possible that massage might have a similar effect. Abscesses or their treatment may precipitate paralysis. The median age of paralysis fell by almost a half from about 2 yr to 1 yr, but may now have risen as many younger children receive vaccine. Rehabilitation has been neglected, with long lasting consequences. Assessment of disability should be based on need and not on current ability. Ethically, prevalence surveys should offer opportunity for immunization and rehabilitation. Prevalence of paralysis, numbers attending for rehabilitation and immunization have been analysed by gender and differences examined. The gender-gap widens with age after paralysis: there may be a high mortality among girls with paralysis. Past prevalence calculated from surveys might be seriously underestimated if many girls have died. The monthly pattern of polio was no different in epidemic years. Pulse immunization could be tried just before the seasonal rise each year. Difficulties of comparing vaccination schedules and the criteria for assessing the potency of vaccine are analysed. Paralysis among Indian soldiers in World War II suggests that adult cases may occur, but are not reported. Present investigations should concentrate on babies and adult cases. Research using virulent strains should be prohibited. The very low case-fatality rate suggests that many of the circulating viruses are of low virulence. The increasing proportion of cases with unnecessary injections just prior to paralysis might have caused the lower median age of paralysis, the severity of paralysis and at least part of the increasing prevalence of polio. The seasonal increases of polio might, in part, be a reflection of the injections given for fever caused by other infections. A national campaign against unnecessary injections for young children is urgently required. For children with fever, there is a strong case for postponing even DPT injections.