The purpose of this study was to develop an in vitro perfusion technique or "continuous-flow adhesion cell" model to predict the in vivo performances of different mucoadhesive drug delivery systems based on hydrogels. Two studies were performed, either using a rabbit small intestine or a polyethylene surface; the adhesion of four gels--two poly(acrylic acid)s (PAAs) (carbomer [CM] and polycarbophil [PC]), an ethyleneoxide-propyleneoxide block copolymer (Poloxamer 407 [PM]), and a polysaccharide (scleroglucane [SG])--were evaluated. In this respect, scleroglucane was used as a control. The adhesiveness of the different gels for both supports is in accordance with that described in the literature, that is, polycarbophil adhered more strongly than carbomer, which itself adhered more strongly than poloxamer. This study proved that the gels adhere more strongly to the polyethylene tube than to the rabbit small intestine, thus indicating that evidence for adhesion properties does not need any presence of mucus. Therefore, our in vitro model could be a good method, more precise and more simple than an ex vivo technique, to predict the bioadhesion of gelified devices.