This is a comment on arXiv:physics/0608292, subsequently published in Phys Rev Lett, 98, 014501 (2007), in which Willis and Kerswell study in direct numerical simulations the statistics of turbulent lifetimes in pipe flow. They suggest that their data for six different Reynolds numbers support a divergence of the lifetime near a Reynolds number of about 1870. However, their conclusion is not compelling: a re-analysis of their data shows that it is also compatible with an exponentially increasing lifetime.