We undertook an in vitro investigation of the role of the design of the eyedrop dispenser in bacterial contamination. The nozzle tips of pipette and squeeze bottles containing Fluress (pH 5.0) were inoculated with 10 microliter of an ocular isolate of Pseudomonas aeruginosa (5.5 X 10(5) bacteria/ml). Cultures of single drops of ophthalmic solution (25-microliter drops from each pipette bottle and 40-microliter drops from each squeeze bottle) were done one minute, 15 minutes, one hour, two hours, and 24 hours after inoculation. Swabs from the inside of the caps of the eyedrop bottles were also cultured at similar intervals. No bacteria were recovered from either dispenser type after one hour. Swabbings from the caps of the pipette bottles showed no growth within minutes after inoculation, but swabbings from the caps of the squeeze bottles consistently yielded bacteria for 24 hours. We suggest that the cap of the squeeze bottle serves as a potential reservoir for bacterial contamination whereas direct contact of microorganisms with the preservative in an ophthalmic solution by the use of a pipette-type dispenser decreases the risk of microbial contamination and growth.