One of the obstacles to chronic GVHD research is the lack of standardized response criteria. The National Institute of Health (NIH) has recommended response criteria at a Consensus Conference. These need to be validated. We recently completed and reported a trial of pentostatin in treating steroid-refractory chronic GVHD. During the trial, we prospectively collected percent body-surface-area (BSA) involvement of rash, superficial sclerosis and deep sclerosis. Here, we compare cutaneous responses using the NIH scale and the Hopkins scale. The two scales produced similar overall response rates but different domain response rates. There was 80% agreement in overall response at the final treatment evaluation, but only a 64% agreement for fasciitis/non-moveable sclerosis. There was more disparity in the measurement of sclerosis than in that of erythema, which highlights the difficulty of quantifying sclerosis. For sclerosis, the Hopkins scale, which used skin softening, was more predictive of early response as compared with the NIH scale, which focused on percent BSA. Early assessment of skin softening may be important if trying to detect the activity of a particular agent in chronic GVHD. Further validation of the NIH scale is ongoing, which should produce a clinically useful and predictive scale.