The clinical tests to diagnose keratoconjunctivitis sicca are weak, whereas the laboratory tests are very powerful, but there is no reason whatsoever to discard the present clinical tests. For example, finding a Schirmer test consistently lower than 5 mm is an easy way to strongly suspect a dry eye state. The level of tear fluid production can help one to determine the preferred viscosity of the tear substitute. Then, what test should be used other than the rose Bengal test to evaluate the over-all effect of the dry eye state on the external eye? What other test than the break-up time is there to evaluate tear film stability? There is no doubt that laboratory tests will be used in the future as routine clinical procedures, in addition to but not replacing the clinical tests, as the tear protein concentration, the Schirmer test values and the break-up time are important for the diagnosis of dry eye and, analyzed together, they may be of value to differentiate between Sjögren's and non-Sjögren's keratoconjunctivitis sicca.