In a study of human Purkinje cell (PC) number, a striking mismatch between the number of PCs observed with the Nissl stain and the number of PCs immunopositive for calbindin-D28k (CB) was identified in 2 of the 10 brains examined. In the remaining eight brains this mismatch was not observed. Further, in these eight brains, analysis of CB immunostained sections counterstained with the Nissl stain revealed that more than 99% Nissl stained PCs were also immunopositive for CB. In contrast, in the two discordant brains, only 10-20% of CB immunopositive PCs were also identified with the Nissl stain. Although this finding was unexpected, a historical survey of the literature revealed that Spielmeyer [Spielmeyer W. Histopathologie des nervensystems. Julius Springer: Berlin; 1922. p. 56-79] described human cases with PCs that lacked the expected Nissl staining intensity, an important historical finding and critical issue when studying postmortem human brains. The reason for this failure in Nissl staining is not entirely clear, but it may result from premortem circumstances since it is not accounted for by postmortem delay or processing variables. Regardless of the exact cause, these observations suggest that Nissl staining may not be a reliable marker for PCs and that CB is an excellent alternative marker.