A general practice population of around 6700 was kept under clinical and laboratory surveillance from 1962 to 1975. Illnesses totalled 18703 in three morbidity classes: sore throat (Throats) 4451, acute febrile respiratory diseases (FRD) 4934, acute non-febrile respiratory diseases (Non-FRD) 9318. Specimens were examined for beta-haemolytic streptococci (BHS) from 37·1% of these illnesses: from Throats 33·3%, from FRD 67·8%, from Non-FRD 22·6%, and 515 specimens were collected from a miscellaneous (‘Other’) class consisting of healthy persons and ailments that could not have had a streptococcal component.

Strains of BHS were isolated from 7448 specimens as follows: group A (Streptococcus pyogenes) 353, group C 36, group G 15, other groups 274.

Group A strains were isolated from specimens at the following rates: Throats 16·7 %, FRD 2·4%, Non-FRD 0·9 %, Other 1·4 %. The last two classes reflect the carrier rate in the general community, which must be deducted to obtain the streptococcal morbidity in the other classes. Carriers thus accounted for 6% of the strains isolated from the Throats class and for 42 % of those from FRD illnesses.

No consistent seasonal trend of prevalence was detected. Long-term fluctuations in prevalence over several years affected all groups and most group A serotypes.

Serotyping was performed on 304 strains from 1963 to 1975. The commonest types found were T-types 4 and 12 and M-type 12.

Immunity against re-infection by identical strains appeared to be fairly strong and also against heterotypic strains that shared a T-antigen, but little protection was conferred against re-infection by group A strains with no shared M- or T-antigen. R-28 antigen is considered here as a marker epidemiologically equivalent to an M-antigen.

Epidemicity, as measured by a simple estimate of aggregation, appeared to be low and there were differences between and within serotypes.

The infecting organism appeared to linger in the pharynx, sometimes for several months, after a streptococcal illness.