Acoustic reflex (AR) threshold and growth were measured in one ear for each of 20 normal-hearing individuals as pressure in the ear canal was varied between k120 mm of H2O. These same parameters were studied in 10 of the subjects after disruption of the probe seal. Activating stimuli for AR threshold measures were 500, 1000, 2000, and 4000 Hz tones and broadband noise. The AR growth function was generated by a 500 Hz tone. As ear canal pressure was increased or decreased relative to the point of maximal compliance, systematic increases in AR threshold were observed for all activating stimuli. These mean pressure effects were relatively small (4 to 8 dB) and were essentially independent of stimulus composition. The ear canal pressure variations had no effect on the relative difference between tone and broadband noise AR thresholds. Absence of an airtight seal had a negligible effect on AR threshold. Growth of the AR was significantly altered by application of a relatively mild positive or negative canal pressure. Absence of a probe seal which had no effect on AR threshold was found to significantly reduce magnitude of the AR. These findings indicate that caution should be exercised during the measurement of AR growth to ensure both proper maintenance of desired canal pressure and canal-probe coupling.