PURPOSE: To examine the role of routine smears, cultures, and antibiotic sensitivity testing in the treatment of suspected infectious keratitis. METHODS: A retrospective chart and laboratory data review was performed for 81 consecutive patients seen in the Los Angeles County/University of Southern California Department of Ophthalmology between June 1991 and December 1993 with a primary diagnosis of community-acquired infectious keratitis. No patients were treated with antibiotics before evaluation in the author's department, and all underwent corneal scrapings for gram-stain and bacterial, fungal, and mycobacterium cultures. Ulcers were classified as moderate or severe. All initially were treated as inpatients with a regimen including fortified cefazolin and a fortified aminoglycoside. RESULTS: Of 81 patients, 74 ulcers were either culture- negative (n=18) or grew bacteria (n=56). Fungal infection was suggested in seven patients. Of the nonfungal ulcers, 33 were classified as moderate, and 41 as severe; all moderate ulcers improved without requiring a modification in antibiotic treatment, whereas 3 severe ulcers required a change in treatment. CONCLUSION: Most community-acquired bacterial ulcers resolve with broad spectrum empiric therapy. Alternatives to universal culture and sensitivity testing that might be considered include selectively performing cultures for more severe or suspected non-bacterial ulcers or routinely obtaining cultures in all cases, but pursuing identification and sensitivity studies only when those data are required for therapy modification.