BACKGROUND: Twenty percent of cholecystectomies in the US are performed for a diagnosis of biliary dyskinesia. Diagnosis is made by measuring gallbladder ejection fraction (GbEF) using hepatobiliary scintigraphy. Our purpose was to evaluate the reproducibility of GbEF measurements. STUDY DESIGN: This is a retrospective review of patients referred for cholecystectomy, from 2010 to 2016, with a diagnosis of biliary dyskinesia based on a GbEF test, who then underwent a repeat GbEF test. Thirty consecutive patients were identified by hospital records. Re-testing of GbEF was performed at least 6 weeks after the initial test using Tc-99m and slow injection of sincalide at 0.02 mcg/kg. RESULTS: On re-testing, 16 of 30 patients (53%) patients had a normal GbEF of >35%, ie the initial test result was not reproducible in them. Age, sex, days between testing, and initial GbEF did not differ between groups. The 14 patients who re-tested positive for biliary dyskinesia with reduced GbEF were significantly more likely to have episodic pain than steady pain. Re-testing frequently resulted in change in management in that most patients who re-tested in the normal range were not offered cholecystectomy. CONCLUSIONS: Hepatobiliary scintigraphy with GbEF is a poorly reproducible test. Re-testing resulted in a change in management in many patients who then avoided cholecystectomy. Strong consideration should be given to repeating hepatobiliary scintigraphy with GbEF before cholecystectomy in patients with an initial positive test.