OBJECTIVE The compatibility of calcium and phosphate salts in total parenteral nutrient (TPN) admixtures at the highest concentrations recommended for preterm and term infants was studied. METHODS Particulate matter from eight different macronutrient combinations was measured and counted (range, 1.8-50 mum) by a laser-based, single-particle optical sensing technique. Measurements were performed at four intervals after compounding the formulations under aseptic conditions (within 1 hour of preparation and at 6, 24, and 30 hours) at 23-27 degrees C. The number of particles measuring >or=5, >or=10, and >or=25 microm per milliliter of TPN admixture was recorded. Detailed visual inspections were also performed at these intervals, and pH was measured at the beginning (time 0) and end of the study (30 hours). Precipitated material was characterized by polarized microscopy and infrared spectroscopy. RESULTS The TPN admixture with the lowest concentration of amino acids (0.5%), as well as the highest pH, resulted in significant growth of particulate matter over time. At 30 hours, the particle growth was accompanied by visible evidence of precipitation, which was confirmed to be dibasic calcium phosphate. Neither significant particle growth nor precipitation was noted in the remaining seven formulations, which had amino acid concentrations of 1-4%. CONCLUSIONS Commonly used organic calcium and inorganic phosphate salts in cysteine-added, lipid-free TPN formulations at the highest recommended amounts for neonates were compatible when the amino acid concentration was between 1% and 4% and the dextrose concentration was 5% or 10%. The salts remained compatible for up to 30 hours at a room temperature of up to 27 degrees C. Precipitation of dibasic calcium phosphate occurred with lower amino acid concentrations and higher pH values.