OBJECTIVES: To determine the role of equine herpesvirus (EHV) in idiopathic keratoconjunctivitis in horses and to determine whether sample collection method affects detection of EHV DNA by quantitative polymerase chain reaction (qPCR). ANIMALS STUDIED: Twelve horses with idiopathic keratoconjunctivitis and six horses without signs of ophthalmic disease. PROCEDURES: Conjunctival swabs, corneal scrapings, and conjunctival biopsies were collected from 18 horses: 12 clinical cases with idiopathic keratoconjunctivitis and six euthanized controls. In horses with both eyes involved, the samples were taken from the eye judged to be more severely affected. Samples were tested with qPCR for EHV‐1, EHV‐2, EHV‐4, and EHV‐5 DNA. Quantity of EHV DNA and viral replicative activity were compared between the two populations and among the different sampling techniques; relative sensitivities of the sampling techniques were determined. RESULTS: Prevalence of EHV DNA as assessed by qPCR did not differ significantly between control horses and those with idiopathic keratoconjunctivitis. Sampling by conjunctival swab was more likely to yield viral DNA as assessed by qPCR than was conjunctival biopsy. EHV‐1 and EHV‐4 DNA were not detected in either normal or IKC‐affected horses; EHV‐2 DNA was detected in two of 12 affected horses but not in normal horses. EHV‐5 DNA was commonly found in ophthalmically normal horses and horses with idiopathic keratoconjunctivitis. CONCLUSIONS: Because EHV‐5 DNA was commonly found in control horses and in horses with idiopathic keratoconjunctivitis, qPCR was not useful for the etiological diagnosis of equine keratoconjunctivitis. Conjunctival swabs were significantly better at obtaining viral DNA samples than conjunctival biopsy in horses in which EHV‐5 DNA was found.