The WHO member countries undertook to eliminate poliomyelitis globally by the year 2000. Acute flaccid paralysis (AFP) surveillance finds paralytic cases of polio so that swift action can be taken and shows that wild poliovirus has been eliminated when polio cases no longer occur. Mpumalanga Province, a rural province in the north-east of South Africa, developed a rapid reporting system where infection control nurses at the public and private sector hospitals report weekly to the AFP surveillance Officer, including zero reporting, on 9 infectious disease syndromes that require rapid action on clinical presentation alone. This system was implemented in 1998 and included AFP. The non-polio AFP reporting rate increased from 0.37 in 1997 to 0.55 during 1998 with more than 80% of the units reporting weekly. The binomial exact confidence intervals however include 1. A hospital record review of all paediatric admissions revealed that only 2 AFP cases were missed by the system. The AFP reporting rate remains below the international standard of 1 per 100 000 children under 15 years of age despite an adequate reporting system. The role of chance variation, particularly in small geographical areas, has not been discussed in official polio-eradication guidelines but it is . imperative that population size be taken into account when judging the rate of AFP case detection. With the low international reference rate and play of chance variation it is possible that regions with relatively small populations, low non-polio AFP detection rates and no cases of polio detected for an extended period may have adequate surveillance systems supporting polio free certification. In these areas additional criteria for determining the adequacy of the surveillance system should also be considered.