The purpose of this comment is to counsel caution in some of the conclusions drawn in an otherwise fine article recently published in Economics and Human Biology on infant mortality in Armenia by Hakobyan, Mkrtchyan and Yepiskoposyan. These relate first, to the reliability of estimates and trends in infant mortality estimated from DHS data; second, to the interpretation of what the authors consider to be a ‘low’ infant mortality rate in former communist countries given their level of economic development; and third, to the role of the health care infrastructure in countries of the former Soviet Union in producing these ‘low’ infant mortality levels. This comment argues that trends in infant mortality in Armenia and other CIS countries, although probably declining, are perhaps less certain than the authors allow, that existing evidence does not suggest that they are uniformly low by global standards, or that the health care systems in CIS countries are uniformly effective in reducing infant deaths.

Keywords: Infant mortality; Transition countries; Survey data; Health care systems