The maximum solubilities of calcium and phosphate in neonatal total parenteral nutrient (TPN) solutions compounded using Aminosyn-PF or TrophAmine amino acid injection were determined. Eight solutions were compounded from Sterile Water for Injection, USP, 50% dextrose injection, usual electrolytes and trace metals, and Aminosyn-PF 7% or TrophAmine 6% to obtain concentrations of amino acids 2.5% and dextrose 25% and amino acids 1% and dextrose 10%. Cysteine hydrochloride was added to half of the solutions at a concentration of 40 mg/g of protein. The pH of each solution was determined at various times during admixture preparation. The solutions were divided into 20-mL aliquots, and phosphate and calcium were added to each aliquot at concentrations of 2.5 to 50 meq/L of calcium and 5 to 50 mmol/L of phosphate. All solutions were inspected visually for signs of precipitation or crystallization after 18 hours of storage at room temperature and again after 30 minutes in a 37 +/- 1 degrees C water bath. Samples with no visible signs of precipitation were filtered, and the filter membrane was inspected under a microscope. Solubility curves were prepared by plotting the concentrations of calcium and phosphate at which either visual or microscopic precipitation occurred. Calcium and phosphate solubility was greater in the solutions with higher concentrations of amino acids and dextrose. In contrast to the results of some previous studies, no important differences in calcium and phosphate solubility were observed between Aminosyn-PF and TrophAmine in solutions containing amino acid concentrations of 2.5%, with or without added cysteine hydrochloride.(ABSTRACT TRUNCATED AT 250 WORDS)