At neutral pH, metal radionuclides, e.g., In-111 and Y-90, precipitate and adhere to container walls, yielding nonuniform activity distributions in compartments of plastic phantoms used for dosimetry validation. For various concentrations of a chelator, EDTA, we evaluated the uniformity of activity distributions in phantoms, as well as the accuracy of a fluorescent liquid dosimeter, coumarin-3-carboxylic acid (CCA). Solutions of In-111 and Y-90 with activity concentrations of 10-25μCi/cc containing increasing EDTA concentrations (0-25μM) were stored in polymethyl methacrylate (PMMA) phantoms for 24 h to accumulate dose. Activity in the solution was then measured in a gamma counter and/or by μSPECT imaging. Phantoms containing 0.1 mM CCA with EDTA were also irradiated either with a Cs-137 irradiator or with Y-90 to evaluate the effect of EDTA concentration on fluorescence intensity (FI). Activity measurements from images and gamma counting indicated that 40-50% of the activity of In-111 and Y-90 (25μCi/cc) accumulated on phantom walls for EDTA concentration <2.5μM, whereas no adherence was observed for EDTA concentration >2.5μM. For a fixed absorbed dose, fluorescence intensity decreased by <15% when using 2.5μM EDTA. For all EDTA concentrations, a linear dose-FI relationship was observed up to ∼15Gy. For phantom dose measurements using metal radionuclides, solutions should contain ∼1,000 EDTA molecules for each radiolabeled atom to obtain a uniform activity distribution.