<h3>To the Editor.—</h3> A recent publication by Zhu and coworkers<sup>1</sup>examined antibody-positive serum samples from 100 sequential patients on sodium chloride-split skin and by Western immunoblot to distinguish bullous pemphigoid from epidermolysis bullosa acquisita. Five percent of unselected patients were shown to have epidermolysis bullosa acquisita. An editorial in the same issue of theArchives<sup>2</sup>brings the problem of such studies into focus. Less than 50% of patients with epidermolysis bullosa acquisita or bullous pemphigoid have free, circulating antibodies capable of reacting with normal salt-split skin. Thus, a determination by indirect immunofluorescence can only be ascertained in less than half of such individuals. The editorial cites an article by Gammon et al<sup>3</sup>that showed direct immunofluorescence on perilesional skin incubated in 1 mol/L of sodium chloride induces a separation at the dermoepidermal junctionthrough the lamina lucida. This allowed detection of complement or antibody to (1) epidermis alone