As part of the national lipid screening project 927 people with a plasma cholesterol level greater than 6.5 mM were detected by screening 4006 men and women aged 25-59 years. Three years later 801 of the 878 patients eligible for a follow-up study (91%) had been followed up at least once. The median number of follow-up visits was two. The bulk of the workload fell on the nursing staff. The mean decrease in cholesterol level was 8-14% in those receiving dietary advice only, 15-25% in those receiving additional drug treatment and 12% for all patients. A proportion of this decrease must be attributable to regression to the mean, loss to follow up when patients were doing well, and the patients' knowledge of their follow-up date. Data on a group of patients not attending for regular follow up suggest that regression to the mean could account for up to 7% of the cholesterol reduction observed. Screening for hyperlipidaemia in general practice is feasible when the necessary infrastructure is provided, but even with a fairly conservative protocol 3% of those screened received drug treatment.