Rett syndrome (RS), a complex and severely disabling neuromotor disorder affecting young girls, was not internationally recognized until in the middle eighties. In Austria and Sweden only, there are clinical experiences since the sixties. The aims of this study were to 1) define the nucleus group of RS in Sweden, 2) develop a frame of defined stages for describing the clinical progress of disease, 3) identify early developmental patterns and their predictivity, 4) give the natural history of neuromotor impairments and disability in adults, 5) analyse the period of clinical regression as to developmental profiles, patterns and progression of symptoms, 6) search informative neurophysiologic correlates to clinical symptoms and signs. By a 6 year (1984-90), nationwide search procedure, comprising an information and tracing program, a travelling consultant program, and a neuropediatric and laboratory examination program, 88 females were added to those 17, observed in Sweden during the sixties and seventies. By March, 1990, 105 females, aged 19 months-44 years (median 17 1/2 years), had entered this series, which was considered representative for the nucleus group of RS in Sweden. 97/105 females (92.3%) fulfilled the criteria for classic RS, according to internationally accepted clinical criteria, and 8 (7.7%), represented so called "formes frustes". The prevalence for classic RS in the south-west region of Sweden was 1:12.000-13.000 girls, the birth years 1965-76 and 1960-84. Based on 29 females with RS, aged 13-28 years, a clinical staging system was constructed and evaluated in 1985, emphasizing the main different phases of the disorder throughout life. This staging system was found a valuable tool for clinical assessment, documentation, and research approach, and has become internationally accepted and widely used. Further experiences have given rise to minor adjustments, which are proposed. By scrutinizing early documentation from 10 girls, at referral aged 20 months-6 1/4 years (median 3 1/4 years) and retrospectively applying the Denver Developmental Screening Test frames, early achievements, signs and symptoms were reconstructed in 1986. Although characteristic neuromotor and neurobehavioural patterns of help for early suspicion, and even subtle signs, possibly reflecting a disorder already during the first months of life, could be identified, predictive clinical manifestations of diagnostic value were not found.(ABSTRACT TRUNCATED AT 400 WORDS)