Hyaluronate-binding protein (HABP) has been extracted in detergent from the membranes of simian virus 40-transformed 3T3 (SV-3T3) cells (Underhill et al, J Biol Chem 258:8086-8091, 1983). When SV-3T3 cells were treated with trypsin prior to isolation and dissolution of the membranes, no hyaluronate-binding activity could be detected. This indicates that all of the detectable HABP of SV-3T3 cells is located on the external surface of the plasma membrane rather than on internal membranes, which would be inaccessible to the trypsin. The detergent-extracted HABP from SV-3T3 membranes was reconstituted into the membrane of lipid vesicles, which were formed by addition of exogenous phosphatidylcholine and cholic acid to the extracts followed by removal of detergent by dialysis against 0.02 M Tris pH 8.0 in the presence of protease inhibitors. Reconstitution was assessed by sedimentation in a discontinuous sucrose gradient and by gel filtration on Sepharose 4B in the presence and absence of detergent. The characteristics of binding of hyaluronate to the reconstituted HABP were then compared with those studied previously for the original membrane-bound HABP and the detergent-extracted HABP (Underhill et al, J Biol Chem 258:8086-8091, 1983). It was observed previously that binding of hyaluronate to HABP in the cell membranes was of higher affinity and specificity than to HABP in the detergent extracts of these membranes. It was found here that reconstitution of the extracted HABP into the membranes of lipid vesicles led to restoration of affinity of binding to the level observed in the original cell membranes. However, whereas chondroitin sulfate does not compete significantly for binding of hyaluronate to cell membrane-bound HABP, partial competition was observed for the reconstituted HABP as well as for detergent-extracted HABP. Thus, it is concluded that the high affinity of binding of hyaluronate to the plasma membrane of SV-3T3 cells is in part dependent on insertion of the HABP in the membrane, but that other interactions, not duplicated in our reconstitution experiments, must be necessary for the specificity of the HABP.