In order to resolve a discrepancy in the literature, we have examined the in vitro inactivation of human alpha 1-proteinase inhibitor by direct exposures either to whole cigarette smoke or to filtered (i.e., gas-phase) smoke. Wyss and coworkers (2) reported that whole smoke does not inactivate the protein, whereas we reported that gas-phase smoke does. We now find that direct exposure to gas-phase cigarette smoke causes a slightly greater inactivation of the protein than does direct exposure to whole cigarette smoke, confirming our earlier suggestion that whole smoke is less oxidizing than is gas-phase smoke. This difference, however, does not explain the dramatic difference between our previous findings and those of Wyss and coworkers (2). The explanation for the discrepancy lies in the nature of the buffers used. Wyss and coworkers used Tris buffer and the use of Tris quenches the inactivation process almost completely. Our experiments used phosphate buffer. We suggest that Tris is an unsuitable buffer for use in experiments that probe the effects of cigarette smoke.