The HA2 subunit of influenza hemagglutinin is responsible for fusion of the viral and host-cell membranes during infection. An N-terminal 127 amino acid construct of HA2, FHA2-127, is shown to induce lipid mixing of large unilamellar vesicles under endosomal low pH conditions. Thus, FHA2 could serve as a good model system for biophysical studies of membrane fusion. With FHA2, we began to develop a mechanistic model which could explain how this short construct facilitates membrane fusion. In this endeavor, we studied the possible role of the kinked loop region (amino acids 105-113). A construct missing this loop, FHA2-90, although able to induce lipid mixing, has lost the sharp pH-dependent transition seen with FHA2-127 and native HA. In addition, FHA2-127 promotes extensive vesicle aggregation more effectively than FHA2-90 upon acidification. These data suggest that the kinked loop may play a pH-dependent regulatory role. To test this, we compared bis-ANS binding to the two constructs and observed that binding to FHA2-127 increases at a faster rate than FHA2-90 as the pH is decreased, indicating that the kinked loop not only is an ANS-binding site, but that it binds better at low pH. The pH dependence of this transition directly correlates with that observed in lipid mixing. Further, cysteine mutations of acidic residues in the kinked region are both fusion inactive and bind much less ANS, whereas a similar mutation of a threonine residue had little effect on fusion activity or ANS binding. This evidence lends further support to our idea that the kinked loop serves a regulatory role. To test the physiological relevance of the FHA2-127 fusion mechanism, we studied the effects of a G1E mutation, known to abolish fusion in native HA. We found that G1E-127 is fusion inactive as expected. This evidence indirectly suggests that the mechanism of FHA2-127 is perhaps physiologically relevant and from its study, we can learn much about the mechanism of native HA.