Improvements in the current test (GSR2003) were as follows: three-component particles with four different sizes, including sub-mm-particles, were prepared; a thin carbon layer was applied on the silicon substrate that approached "real" GSR samples by using a carbon-containing adhesive tape; and an extensive statistical evaluation of the obtained data was conducted by independent statisticians. This report's description of the materials and methods addresses the preparation of test material, the homogeneity and stability of the test material, the organization of the proficiency test, the laboratory evaluation and data assessment, and the estimation of the method's detection capability. The new sample design facilitates an estimation of the overall method's detection capability. Detection capability was found to be a function of the particle size. As expected, the method's detection capability decreased with a decrease in particle size. To achieve a detection capability of at least 90 percent, the particle size must exceed 1.8 mm. For 1.0 mm particles, the detection capability is only 65 percent at present. The aim of future proficiency tests will be a further optimization of the parameter setting in order to enhance the sensitivity of the method and to harmonize the performance of the GSR investigation in various laboratories. 9 tables, 3 figures, and 17 references