To determine the reliability of assessment of alpha-synuclein-immunoreactive (alphaS-IR) structures by neuropathologists, 28 evaluators from 17 centers of BrainNet Europe examined current methods and reproducibility of alphaS-IR evaluation using a tissue microarray (TMA) technique. Tissue microarray blocks were constructed of samples from the participating centers that contained alphaS-IR structures. Slides from these blocks were stained in each center and assessed for neuronal perikaryal inclusions, neurites, and glial cytoplasmic inclusions. The study was performed in 2 phases. First, the TMA slides were stained with the antibody of the center's choice. In this phase, 59% of the sections were of good or acceptable quality, and 4 of 9 antibodies used performed consistently. Differences in interpretation and categorization of alphaS-IR structures, however, led to differing results between the laboratories. Prior to the second phase, the neuropathologists participated in a training session on the evaluation of alphaS-IR structures. Based on the results of the first phase, selected antibodies using designated antigen retrieval methods were then applied to TMA slides in the second phase. When the designated methods of both staining and evaluation were applied, all 26 subsequently stained TMA sections evaluated were of good/acceptable quality, and a high level of concordance in the assessment of the presence or absence of specific alphaS-IR structures was achieved. A semiquantitative assessment of alphaS-IR neuronal perikaryal inclusions yielded agreements ranging from 49% to 82%, with best concordance in cortical core samples. These results suggest that rigorous methodology and dichotomized assessment (i.e. determining the presence or absence of alphaS-IR) should be applied, and that semiquantitative assessment can be recommended only for the cortical samples. Moreover, the study demonstrates that there are limitations in the scoring of alphaS-IR structures.