Five adult cats were given 100 mg chloramphenicolper oson five occasions: twice in tablet form, twice in capsules which also contained barium sulphate, and once in capsules without barium. Plasma samples for chloramphenicol assay were obtained on either a ‘standard’ schedule (four samples collected within 8 h of dosing with tablet, plain capsule or capsule including barium) or a ‘frequent’ schedule (nine samples collected within 8 h of dosing with tablet or capsule including barium). With frequent sampling, abdominal radiographs were taken at intervals after administration of capsules. No differences were found with the frequent schedule between tablet and capsule with regard to antibiotic concentrations at each sampling, peak concentrations, or area under the curve of plasma concentration versus time. Standard sampling gave similar drug concentrations to those obtained with frequent sampling. Delayed absorption (defined as plasma chloramphenicol concentration<5 μg/ml for 1.5 h after dosing) occurred after six of the twenty‐five administrations: it was not related to the frequency of sample collection or the use of capsules rather than tablets, but occurred mainly in certain cats. Poor absorption was shown radiographically in one cat to be associated with delayed dispersal of capsular contents