In 8 patients with psoriasis vulgaris, the cutaneous blood flow (CBF) was measured simultaneously in both involved and uninvolved psoriatic skin before (i.e., on the first day of hospitalization) and on the 3rd, 7th, 14th, and 28th days of treatment with tar. The 133Xe washout method was used after epicutaneous labeling and compared to the laser Doppler velocimetry (LDV) technique. Control experiments were performed in 10 normal individuals. Before treatment the mean CBF in involved psoriatic skin was 62.6 +/- 18.7 SD ml X (100 g X min)-1, which is significantly higher than CBF of uninvolved skin in psoriatic patients, 9.5 +/- 4.0 SD ml X (100 g X min)-1, (p less than 0.01) and is 13.6 times higher than CBF in the normal individuals (p less than 0.01). Fifty hours following onset of treatment (i.e., after only 2 applications of tar), mean CBF of the involved psoriatic skin had decreased significantly to 35.0 +/- 13.9 SD ml X (100 g X min)-1, (p less than 0.01), which was not statistically different from the CBF on the 7th day. During the following weeks, the CBF in involved psoriatic skin decreased at a more moderate rate than that observed during the first week and was 15.0 +/- 6.1 SD ml X (100 g X min)-1 on the 28th day. This value is not significantly different from the CBF of uninvolved skin in these patients. At the end of treatment, the CBF of the uninvolved skin had decreased significantly (p less than 0.05) in all the patients to values similar to those observed in the skin of normal individuals. A parallel decline was observed in a clinical psoriatic score index; however, it is not known whether the observed decrease in CBF was preceded or succeeded by the clinical improvement. A comparison of the 133Xe measurements and LDV measurements in the normal individuals by linear regression analysis yielded a correlation coefficient of -0.24 (N = 20, p greater than 0.05). In the skin of psoriatic patients the correlation coefficient was 0.01, (N = 47, p greater than 0.05) for unaffected skin, and 0.61 (N = 47, p less than 0.001) in the involved psoriatic skin sites. The LDV measurements did not reflect changes in the uninvolved skin in psoriatic patients during treatment and resulted in a remarkably high C.V. in the bilateral measurements of skin in normal individuals (43.8%) compared to the C.V. of 133Xe flow measurements (14.8%). It is concluded that while the LDV method gives a rough estimate of blood flow in cutaneous tissue with a high capillary perfusion rate, inaccurate measurements are made in skin areas with normal to twice the normal CBF range.