The present study is concerned with artifacts likely to occur in a horseradish peroxidase exclusion test. Incubation of murine peritoneal macrophages and lymphocytes with the peroxidase showed a close relationship between the number of living cells and the percentage of cells excluding the tracer. The penetration of the cytoplasm by horseradish peroxidase is attributed to an increase in the permeability of the cell membrane during the incubation (ranging from 10 to 120 min). It was not increased by the presence of tracer throughout the incubation period. However, concomitant fixation of the cell in the presence of horseradish peroxidase caused an increase in the influx of the tracer. The horseradish peroxidase exclusing test applied to the guinea-pig organ of Corti has proved to be valid provided that: (a) mechanical lesions prior to the tracer incubation are avoided; (b) incubation is terminated by removal of the extracellular tracer; (c) fixation is carried out as soon as possible; (d) a low concentration of horseradish peroxidase is used; and (e) specimens are incubated in diaminobenzidine-H2O2 medium for the shortest possible period. Although fixation-induced cytoplasmic infiltration by horseradish peroxidase was not detected in cochlear specimens, the findings call attention to possible sources of error and define the level of significance of the test. Horseradish peroxidase does not appear to be a cytotoxic agent under the conditions used.