Abstract
Using the census data for Peru, Bolivia and Ecuador, previous writers have investigated some possible determinants of inter-regional differences in fertility; language spoken, female participation rates, and altitude. This paper points out the many sources of inaccuracy in the census data used. It argues that the indicators of unusually low fertility in the highland, predominantly Indianspeaking areas fail to control effectively for the very high levels of infant mortality in these regions. Fragmentary survey results give some indication of the scale of infant mortality, and appear to refute the idea that fertility is exceptionally low in areas of high altitude. In an attempt to explain why such high mortality rates persist in the Andean region the main health problems of Bolivia are examined. It seems that the causes are economic and social, rather than physiological. Unfortunately a change of policy which reduced death rates would produce grave new social problems.