Stereological quantification of mast cell numbers was applied to sections of punch biopsies from lesional and nonlesional skin of atopic dermatitis patients and skin of healthy volunteers. We also investigated whether the method of staining and/or the fixative influenced the results of the determination of the mast cell profile numbers. The punch biopsies were taken from the same four locations in both atopic dermatitis patients and normal individuals. The locations were the scalp, neck and flexure of the elbow (lesional skin), and nates (nonlesional skin). Clinical scoring was carried out at the site of each biopsy. After fixation and plastic embedding, the biopsies were cut into 2 microns serial sections. Ten sections, 30 microns apart, from each biopsy were examined and stained alternately with either toluidine blue or Giemsa stain and mast cell profile numbers were determined. The study yielded the following results: (1) in atopic dermatitis lesional skin an increased number of mast cell profiles was found as compared with nonlesional skin, (2) comparing atopic dermatitis skin with normal skin, a significantly increased number of mast cell profiles per millimetre squared was found in specimens from the neck, (3) staining with toluidine blue yielded a lower number of mast cell profiles than Giemsa staining, (4) the use of Carnoy's fixative resulted in a lower mast cell profile count than the use of formaldehyde, and (5) there was no statistically significant correlation between the clinical score and the number of mast cell profiles per millimetre squared. Using stereological techniques, this study indicated that mast cells might participate in the inflammatory process in skin leading to atopic dermatitis.