A longitudinal study of patients with persistent hyperamylasemia was carried out to evaluate the clinical significance of this condition. Twenty-five outpatients were studied by means of serum amylase, isoamylase (wheat germ-inhibition method), and lipase determination; macroamylase detection; and abdominal ultrasonography over a one-year period. Cellulose acetate electrophoresis was carried out to validate the wheat germ-inhibition tests; the results of the two assays were closely correlated, except in three patients with macroamylasemia. At the time of the study, none of the patients had evident signs or symptoms of pancreatic disease. At initial evaluation, wheat germ test demonstrated an elevation of only salivary isoamylase in 16 patients, both pancreatic and salivary isoenzyme in two, and only pancreatic isoamylase in six patients (three with macroamylasemia). Normal salivary and pancreatic isoenzymes were found in one with predominantly pancreatic isoamylase. At the 12-month follow-up, hyperamylasemia had disappeared in six cases and salivary isoamylase elevation in three; pancreatic isoamylase remained abnormally high in all eight patients in whom it was elevated at initial evaluation, and lipase was abnormally high in three patients with elevated pancreatic isoamylase. Of the five patients with true pancreatic hyperamylasemia, one had a juxtapapillary duodenal diverticulum, one showed a slight ultrasound alteration of the pancreas, and one had a past history of acute pancreatitis. In our study, most cases of chronic hyperamylasemia were of nonpancreatic origin. In the patients with elevated pancreatic isoamylase, there was no clinical evidence of pancreatic damage, although a subclinical pancreatic involvement could not be excluded in some.