Sultamicillin (CAS 76497-13-7) is a prodrug combination of ampicillin (CAS 69-53-4) and sulbactam (CAS 68373-14-8), with the antibiotic ampicillin and the beta-lactamase inhibitor sulbactam chemically linked as double ester. The present study was performed to investigate the relative bioavailability and to assess the bioequivalence of two different sultamicillin suspensions (Devasid (R) 250 mg/5 ml as test preparation and 375 mg/7.5 ml of the originator product as reference preparation). Twenty-four healthy male volunteers received equal doses of the sultamicillin preparations according to an open, randomised, single-dose, two-period cross-over design with a wash-out phase of 7 days. Blood samples for pharmacokinetic profiling were taken up to 8 h post-dose, and ampicillin and sulbactam plasma concentrations were determined with a validated LC-MS/MS method. Maximum plasma concentrations (C-max) of 11,267.4 ng/ml (ampicillin, test), 10,864.4 ng/ml (ampicillin, reference), 6,360.6 ng/ml (sulbactam, test and 6,410.7 ng/ml (sulbactam, reference) were achieved. Areas under the plasma concentration-time curve (AUC(0-infinity)) of 17,512.9 ng . h/ml (ampicillin, test), 18,388.0 ng . h/ml (ampicillin, reference), 10,971.7 ng . h/ml (sulbactam, test) and 11,181.2 ng . h/ml (sulbactam, reference) were calculated. The median t(max) was 0.69 h (ampicillin, test), 0.83 h (ampicillin, reference), 0.72 h (sulbactam, Devasid (R)) and 0.83 h (sulbactam, reference). Plasma elimination half-lives (t(1/2)) of 1.04 h (ampicillin, test), 1.03 h (ampicillin, reference), 1.26 h (sulbactam, Devasid (R)) and 1.00 h (sulbactam, reference) were determined. Both primary target parameters AUC(0-infinity) and C-max of ampicillin and sulbactam were tested parametrically by analysis of variance (ANOVA) and the 90% confidence intervals were between 84.58%-117.80% (AUC(0-infinity), ampicillin), 92.37%-119.93% (C-max, ampicillin), 85.81%-120.50% (AUC(0-infinity) sulbactam) and 88.41%-117.57% (C-max, sulbactam). Bioequivalence between test and reference preparation was demonstrated since for both parameters AUC and C-max the 90% confidence intervals of the T/R-ratios of logarithmically transformed data were in the generally accepted range of 80%-125%.