Electrostatic and hydrophobic forces are generally recognized as important in bacterial adhesion. Current continuum models for these forces often wrongly predict measurements of bacterial adhesion forces. The hypothesis tested here is that even qualitative guides to bacterial adhesion often require more than continuum information about hydrophobic forces; they require knowledge about molecular details of the bacteria and substrate surface. In this study, four different strains of bacteria were adsorbed to silica surfaces hydrophobized with alkylsilanes. The thickness of the lipopolysaccharide layers varied on the different bacteria, and the lengths of the alkylsilane molecules were varied from experiment to experiment. Bacterial adhesion was assessed using column experiments and atomic force microscopy (AFM) experiments. Results show that hydrophobized surfaces have higher bacterial sticking coefficients and stronger adhesion forces than bare silica surfaces, as expected. However, adhesion decreased as the solution Debye length became longer than the alkylsilane, perhaps since the silane molecules could not "reach" the bacterial surface. Similarly, those bacteria with a long o-antigen layer had decreased adhesion, perhaps since the silane molecules could not reach surface-bound proteins on the bacteria. This study reveals that macroscopic measurements such as contact angle are not able to fully describe bacterial adhesion; rather, additional details such as the molecular length are required to predict adhesion.