PURPOSE: To evaluate 2 local dilution protocols to assess the accuracy and variability of intracameral antibiotic dosage in cataract surgery. SETTING: Tennent Institute of Ophthalmology, Glasgow, United Kingdom. METHODS: Ten ophthalmic operating room nurses from 2 local hospitals participated. Oven-dried analytical grade potassium chloride (KCl) was used as a surrogate for cefuroxime. Solutions intended for intracameral use (1.0 mg in 0.1 mL) were prepared according to the 2 protocols. Twenty samples were obtained for each protocol. Ten analytical chemists also performed both dilutions. Concentrations of KCl in each 0.1 mL sample were analyzed by flame photometry. RESULTS: Thirty samples were obtained for each protocol. The median dose after dilution was 1.17 mg (range 0.62 to 1.77 mg) for protocol 1 and 2.05 mg (range 0.52 to 7.25 mg) for protocol 2. The median was significantly higher for protocol 2 (P < .001). There was also greater variability with protocol 2. CONCLUSIONS: This study shows that the mathematical accuracy of a dilution protocol does not ensure dosage accuracy in the clinical scenario. Inadequate mixing in a 1.0 mL syringe was probably responsible for the inaccuracy of protocol 2, indicating that small-volume syringes should not be used for mixing. However, protocol 1 had an acceptable range of variability. Replication of this study could evaluate other protocols and address concerns regarding the accuracy of intracameral antibiotic preparations.