Previous studies have shown that the relationship between gastrointestinal symptoms and gastric emptying is weak. Therefore the quantitative assessment of gastric emptying with a relatively simple, non-invasive test would be of considerable clinical value in insulin-treated diabetic patients to identify those with disturbed gastric emptying. The aim of this investigation was to evaluate the inter- and intra-subject variability of a paracetamol-pasta test in healthy subjects and in IDDM patients. Eighteen healthy subjects (8 women) with a mean age of 37 years (range 19-68) and 19 IDDM patients (10 women) with a mean age of 48 years (range 25-62) and mean duration of diabetes of 28 years (range 6-52) were studied on two occasions with an interval of 1 to 4 weeks. After an overnight fast the subjects ingested a standardized pasta meal mixed with 2 g paracetamol in a period of 15 min. Blood samples were drawn at regular intervals after meal intake and analysed for paracetamol (P) and blood glucose. The serum levels of P were significantly lower at 15 min in diabetic patients. The intra-subject coefficients of variation (CV%) of the areas under the serum paracetamol concentration-time curve (AUC) were almost identical in healthy and diabetic subjects, while the intra-subject CV of the P-Tmax was considerably lower in diabetic patients as well as markedly lower than the corresponding inter-subject CV. The inter-subject CVs of all parameters calculated were generally higher in diabetic patients. This study indicates that the assessment of paracetamol absorption kinetics during a paracetamol-pasta test is reproducible in healthy as well as in diabetic subjects. Diabetic patients with non-optimal glucose control and without a case history indicating gastroduodenal motor function disturbances achieve lower serum concentration of P at 15 min and generally display a higher inter-individual variability indicative of subclinical disturbances of gastric emptying in this group of patients.