Hemiplegic cerebral palsy (CP) was studied in a retrospective population-based series of 169 cases from the South-western Swedish health care region covering the birth years 1969-78. The purpose was to analyse the prevalence, aetiology and neuro-developmental outcome in children born preterm and at term, and to correlate pathogenetic periods, aetiological factors and clinical parameters to neuroradiology. The prevalence at the ages 6-15 years was 0.66 per 1000. Postnatally acquired hemiplegia, mainly postinfectious, iatrogenic or posttraumatic, constituted 11\%. Among term children with congenital hemiplegia (pre and perinatally derived) the aetiology was considered prenatal, mainly circulatory brain lesions and maldevelopments, in 42\%, combined pre and perinatal in 9\%, perinatal (cerebral haemorrhage, hypoxia) in 16\% and untraceable in 34\%. The corresponding distribution among preterm children was 29\%, 47\%, 25\% and 6\%, respectively. The rate of preterm birth among congenital cases was 24\%. Birth asphyxia was shown to be a poor indicator of pathogenetic period, whereas a cascade of postpartum complications suggested perinatal brain damage. Clinical follow-up of 152 children revealed that 50\% had mild, 31\% moderate and 19\% severe motor dysfunction. Stereognostic sense was impaired in 44\% of the children (astereognosia in 20\%). Additional impairments (mental retardation, epilepsy, impaired vision, hearing and speech, severe behavioural/perceptual problems) were present in 42\%. Term children with congenital hemiplegia tended to be more severely affected than preterm children. The resulting total handicap was considered mild in 40\%, moderate in 44\% and severe in 16\%. The prevalence of severe total handicap was highest among postnatal cases. Computerised tomography (CT), performed in 109 congenital cases, was normal in 26\%, showed unilateral ventricular enlargement in 36\% and revealed cortical/subcortical cavities in 20\%. In the remaining 18\% CT findings were classified as öther". With the classification so far used, correlations between CT findings and aetiologies were unsatisfactory and disappointing. In contrast, CT findings showed a strong correlation with clinical degree of severity and magnitude of associated handicap. As a rule, normal CT implied mild disability and unilateral ventricular enlargement moderate, whereas cortical/subcortical cavities were frequently associated with severe handicap, including mental retardation and epilepsy.