‘Time and thyme are not homophones’, a 2008 article by Susanne Gahl published in Language, reports a frequency effect differentiating the durations of homophones, for example, time vs. thyme. The article is of fundamental theoretical relevance, as the finding reported has significant implications for research on homophones and the effects of frequency in general. As I show in the present paper, however, the main analysis in Gahl 2008 does not provide quantitative evidence for the effect. The same is true of a follow-up study (Gahl 2009). I provide here a reanalysis based on the original data set, which shows that the frequency effect reported in the original article is real.