Particulate contamination levels of a base intravenous solution and five admixtures were assessed in a double-blind study conducted to determine the effect of inline i.v. filtration on infusion phlebitis. Only a qualitative evaluation on actual patient filters was possible due to deformation of the filter membrane after 72 hours; simulation of i.v. therapy in the laboratory gave more reliable information about particle counts and distribution. The base solution was dextrose 5% in 0.2% sodium chloride injection; the admixtures contained this base solution and the following additives (per liter): (1) 20 mEq potassium chloride, (2) 20 mEq potassium chloride and 2 ml vitamin B complex, (3) 20 mEq potassium chloride and 5 million units of potassium penicillin G, (4) 20 mEq potassium chloride and 1 g cephalothin sodium (delivered in a 100-ml volume from a medication chamber), and (5) 30 mEq potassium chloride and 3 million units of penicillin G (delivered every four hours from a medication chamber) and 80 mg of gentamicin sulfate (delivered every eight hours from a medication chamber). None of the six i.v. infusion schedules simulated exceeded the standards for particulate matter prescribed by the USP XIX. As expected, however, the particle counts increased as a function of the additives used in the simulation, particularly with the antibiotics.