We studied silicone elastomer lens binding in vitro to determine what factors may influence its development on the cornea or corneosclera. Lens binding to corneas was not influenced by corneal toricity (0-20 D), corneal fitting relationship (2 D steep to 4 D flat), mucin (2 or 5%) in the tear-bath, or transcorneal pressure (11-22 mm/Hg). In isolated corneas or in whole eyes, transient intraocular pressure changes did not influence keratometry readings, ruling these out as potential mechanisms for corneal binding during sleep. Corneoscleral preparations were also examined to simulate a decentered lens. Corneoscleral binding occurred with a significantly (P less than 0.001) greater frequency than corneal binding and was not influenced by corneal toricity, corneal fitting relationship (up to .5 mm steeper than K), or mucin concentration. Unlike the final stages of clinical lens adhesion, the binding we observed permitted lateral lens movement and occurred without leaving an indentation ring. These findings may suggest that the system models the initiation of corneoscleral binding, involving decentration and suction onto the corneoscleral junction. Corneal binding, however, cannot be explained by a chemical attraction between the silicone elastomer lens surface and cornea, with or without mucin interaction, and must be accounted for by other factors found in vivo.