Convergently evolved mole-rats (Mammalia, Rodentia) provide a fascinating model for studying convergent molecular evolution. Three genome sequences have recently been made available for the blind mole-rat (Nannospalax galili; Spalacidae; Muroidea)1, and the convergently evolved naked mole-rat (Heterocephalus glaber; Heterocephalidae; Ctenohystrica)2 and its close relative the Damaraland mole-rat (Fukomys damarensis; Bathyergidae; Ctenohystrica)3. In their genome paper1, Fang et al. evaluated convergent molecular evolution related to the subterranean life-style between the naked mole-rat and the blind mole-rat. One particularly striking result was the strong signal for amino acid convergence detected in the circadian rhythm CLOCK gene. Here I show that this unexpected result is erroneous because it is based on the use of the wrong sequence for the naked mole-rat, which has been mistakenly replaced by a sequence from a blind mole-rat. When the correct sequence is used, the evidence for convergent molecular evolution in this gene appears very limited.