Results from endoscopic sphincter of Oddi manometry are being used to support the diagnosis of sphincter dysfunction in patients with unexplained pain after cholecystectomy. However, there are few data on the reproducibility of manometric records or motility diagnosis during a second test. In this study, the reproducibility of manometric records was assessed in 12 patients with pain after cholecystectomy by performing a second study after three months. Manometric tracings were evaluated without access to patients details and scored for sphincter basal pressure, frequency and amplitude of phasic contractions, propagation of phasic contractions, and responses to intravenous injection of cholecystokinin octapeptide (20 ng/kg). At the initial manometric study, four patients were diagnosed as normal, four as stenotic, and four as dyskinetic. Those diagnosed as normal and stenotic at the first study had an identical diagnosis at the second study. However, the diagnosis of dyskinesia was reproduced only in two of the four patients. In the other two patients a diagnosis of "stenosis" and "normal" was made at the second study. Cholecystokinin octapeptide (20 ng/kg intravenous bolus) produced inhibition of phasic contractions in all studies, both initially and at three months. We conclude that endoscopic sphincter of Oddi manometry is reproducible when the initial diagnosis is either normal or stenosis. However, the diagnosis of dyskinesia is poorly reproducible, perhaps due to the episodic nature of this manometric disorder or to progression of sphincter of Oddi dysfunction.