jats:titleAbstract</jats:title>jats:pThe year when total column ozone (TCO) returns to 1980 levels is commonly used to quantify recovery from disturbance caused by ozone‐depleting substances. This date is reasonable but somewhat arbitrary. Here we borrow an indicator from climate change research, the signal‐to‐noise (S/N) metric to investigate how TCO might return to statistically similar levels to those of the pre‐ozone hole era (1960–1970). We calculate S/N for ozone projections from the Chemistry‐Climate Model Initiative. In most regions, a return to 1980 levels doesn't represent TCO recovery to pre‐disturbance conditions because it does not account for internal variability or reflect when statistically significant TCO losses occurred. Future projections show that, in most regions, TCO “de‐emerges” (i.e., S/N becomes less than one standard deviation from the pre‐disturbance mean) before it returns to its 1980 value. We conclude that S/N is an appropriate, perhaps complementary metric for determining when TCO returns to a pre‐disturbance state.</jats:p>