The primary aim of the study was to evaluate practice differences in reported morbidity in the second and third national morbidity surveys (1970/71, 1981/82) and to discuss their cause. A secondary aim concerned the validation of trends identified from analysis of the data from the total populations in the practices. Altogether 19 practices participated in both surveys. Annual prevalences (that is, the number of patients attending the general practitioner with a condition per 1000 persons at risk) were examined for: all conditions; each of three categories of seriousness of disease; diseases aggregated by chapter of the International classification of diseases; and each of 130 rubrics of the disease classification. Annual prevalence for 'all conditions' was approximately the same for males in both surveys, whereas for females there was an increase. In both sexes, annual prevalence for 'serious conditions' increased slightly and for 'trivial conditions' increased substantially. For 'intermediate conditions', there was a modest decrease in males. In the analysis at ICD chapter level, substantial increases in prevalence occurred in infectious diseases, nervous system diseases, circulatory diseases, genitourinary diseases, musculoskeletal diseases, symptoms, signs and ill-defined conditions, injuries and poisonings. Decreases were found in blood diseases, mental disorders and digestive diseases. Among 130 individual conditions examined, increased annual prevalence was found for mumps, fungal infections, hypothyroidism, diabetes, gout, senile dementia, angina, left heart failure, catarrh, hay fever and asthma, orchitis, acne, osteoarthritis and for some symptoms. Decreases were found for iron deficiency anaemia, anxiety state, refractive errors, haemorrhoids, chronic bronchitis, functional disorders of the stomach, carbuncle and skin infections.(ABSTRACT TRUNCATED AT 250 WORDS)