Previously we investigated gastric emptying and distribution of a capsule formulation of microencapsulated KCl and found the drug was usually present in clumps of KCl crystals held in place by gastric mucus. We therefore investigated whether a tablet formulation of microencapsulated KCl would have improved dispersion. We characterized the intragastric disintegration of capsules and tablets of microencapsulated KCl in 12 subjects. The capsule formulation floated in the gastric pool; one end would adhere to the gastric mucosa and the motion of the tethered capsule would pull the end of the capsule off. The KCl crystals would then be deposited in a mass. In contrast, the tablet formulation sank to the anatomically most dependent portion of the stomach. The tablet rapidly became soft and fragile but, if allowed to remain in one place and minimally disturbed, required a median of 12 min to lose its shape. If allowed to reach the gastric antrum, the tablet was quickly ground by the antro-pyloric pump and widely dispersed. Once liberated in the stomach, the microencapsulated KCl crystals were bound into a more-or-less cohesive mass. The differences between KCl formulations, once the crystals were released, was minimal although the larger crystals from the tablet formulation appeared less adherent and cohesive; they dispersed more in a reticulated pattern when the stomach was distended. We conclude that formulation of a drug in a microencapsulated multiple-unit dosage form does not guarantee wide dispersion nor absence of high local concentration of drug.