Esophageal involvement by scleroderma is frequent. Investigation by manometry or radiography is invasive and nonphysiological. Scintigraphy of the clearance of small radiolabelled liquid boluses in the supine position, while sensitive and noninvasive, may also be nonphysiological and does not allow the simultaneous determination of gastric emptying. We thus studied the esophageal clearance of a semisolid test meal ingested in the upright position. Forty-seven patients with scleroderma and 24 with Sjogren's syndrome were compared with ten normal controls and ten patients with gastric emptying abnormalities but no esophageal involvement. Results of scintigraphy were also correlated with manometry and contrast radiography. Quantitative evaluation of esophageal tracer retention at ten minutes postingestion was: (mean +/- SD), 2.8 +/- 1.0% in normals, 2.9 +/- 0.9% in gastric dysmotility, 4.8 +/- 2.9% in Sjogren's syndrome, and 22.3 +/- 25.0% in scleroderma; similar results were found at 20 and 60 minutes. The T 1/2 of gastric emptying was 47.1 +/- 5.7 minutes in normals, 95.9 +/- 25.3 minutes in gastric dysmotility, 62.9 +/- 19.5 minutes in Sjogren's syndrome, and 52.9 +/- 13.5 minutes in scleroderma. We conclude that esophageal clearance of a semisolid test meal is a sensitive index of esophageal dysmotility and correlates well with results from manometry and contrast radiography but is noninvasive and quantifiable. The simultaneous measurement of gastric emptying is also possible in many cases.