Purpose The PCR was compared with routine microbial studies for the detection of fungal pathogens in clinically suspected fungal keratitis. Methods A prospective nonrandomized study was undertaken at a tertiary eye care centre to evaluate 30 eyes of 30 patients with presumed fungal keratitis, both fresh and treated. Corneal scrapings were performed on each patient. The specimens were analysed by a semi-nested PCR assay using fungalspecific primers. PCR products were cloned and sequenced for identification, and compared with a conventional microbial work-up (smear and culture). Results Of the 30 samples, the PCR showed positivity in 93.3%, culture in 40%, and potassium hydroxide in 20%. Of the 28 PCR-positive cases, 12 were culture-positive and 16 were culture-negative. Two samples were both PCR and culture test negative. Culture-negative samples were PCR-positive in 16 of 18 (88.9%) cases. The PCR did not yield any false-negative findings in a culturepositive specimen. Both common and uncommon aetiologic fungi have been identified by DNA sequencing analysis. Conclusion The PCR was able to detect fungal DNA in a high proportion of culturenegative cases. Technical considerations of the PCR process include extraction of artifacts and amplification of non-pathogenic DNA. Nonetheless, our findings suggest that the PCR can be a useful adjunct to smear and culture in the rapid diagnosis of fungal keratitis, particularly in cases of failed detection from routine procedures.