Determination of pancreatic function is essential in cystic fibrosis. The most-reliable method is by measuring pancreatic enzymes in the duodenum following intravenous or oral stimulation. However, this is invasive, time consuming, and expensive. Indirect tests are non-invasive but lack accuracy. This study examines a simple test which combines pancreatic stimulation by Lundh meal and sequential serum lipase measurements. The test was performed on three groups: group A, 36 cystic fibrosis patients carrying two mutations associated with severe disease and pancreatic insufficiency (delta F508, W1282X, G542X, N1303K, S549R); group B, 8 compound heterozygote cystic fibrosis patients carrying one mutation causing mild disease with pancreatic sufficiency (3849 + 10 kb C-->T); group C, 17 healthy individuals. Basal lipase levels were 2-16.5, 16.4-73, and 8.5-27.8 U/l in groups A, B, and C, respectively, with some overlapping between groups. There were three patterns of lipase activity (1) consistently low levels (group A) suggested a severely affected insufficient pancreas; (2) normal basal levels followed by a linear rise peaking 30 min after the meal (found in 16 of 17 healthy individuals and 3 patients of group B) reflecting an unaffected sufficient pancreas; (3) elevated lipase levels not influenced by the meal (5 patients of group B). This reflects an ongoing destructive process in the pancreas which will eventually result in conversion from pancreatic sufficiency to pancreatic insufficiency. Hence serum lipase activity prior to and 30 min after Lundh meal is a good indicator of pancreatic status allowing categorization of cystic fibrosis patients as pancreatic insufficient, pancreatic sufficient, or pancreatic sufficient with late conversion to insufficiency.