Kochanek suggested that the redshifts of gravitational lens galaxies rule out a large cosmological constant. This result was questioned by Helbig & Kayser, who pointed out that selection effects related to the brightness of the lens can bias the results of this test against a high lambda_0 value; however, we did not claim that the observations favoured a high lambda_0 value, merely that current observational data were not sufficient to say either way, using the test as proposed by Kochanek but corrected for selection effects. Kochanek pointed out that an additional observable, namely, the fraction of measured lens redshifts, provides additional information which restores the sensitivity of the test to the cosmological model, at least somewhat. Here, I consider three aspects. First, I discuss the appropriateness of the correction to the test proposed by Kochanek. Second, I briefly mention the slightly different statistical methods which have been used in connection with this test. Third, I discuss what results can be obtained today now that more and better-defined observations are available.