The oral bioavailability of aluminium was compared after administration of 1 g sucralphate as either a tablet or a suspension (1 g/5 ml) in a crossover study in 16 healthy volunteers. Aluminium levels were detectable in all subjects pre‐dose (21.4±8·8 μg 1−1 before tablet; 21·4±7·4 μg 1−1 before suspension) and there was a measurable increase in the plasma concentrations of aluminium in all subjects after administration of the suspension, and in 14 of the subjects after administration of the tablet formulation, with Cmax reached within the first 8 h in most subjects. Plasma levels were still elevated 72 h after dosing. The variability in plasma levels of aluminium was significantly higher after administration of the suspension (CV 39–53%) than after administration of the tablet (CV 29–44%), reflecting greater absorption of aluminium from the suspension formulation in three subjects. Similarly, the variance of the Cmax, AUC0–72 h, and AUC0–∞ (for both the raw data and the baseline adjusted data) were all higher for the suspension than for the tablet. A point estimate of the difference of the pharmacokinetic parameters (determined from the median of the arithmetic Walsh averages) indicated little or no difference in Cmax, Tmax, or AUC0–∞ in the two formulations. In summary, the performance of the suspension formulation of sucralphate is more variable than the tablet formulation in vivo and some patients may therefore have higher circulating levels of aluminium on therapy with the suspension formulation.