The perinatal mortality rate is used as an indicator of the quality of antenatal and perinatal care, yet uncritical application of this indicator in international comparisons can be misleading. The perinatal mortality rate depends on a number of factors and important determinants that need to be assessed separately before reaching conclusions about quality-of-care issues. This article provides a conceptual model of the construction of the perinatal mortality rate. It illustrates the relationship between quality of antenatal and perinatal care and risk factors for perinatal mortality and how these lead to the perinatal mortality rate. It also indicates how differences in registration procedures and practices influence the final mortality figures published by individual countries. For international comparison, the first step is to apply common definitions. The rate can vary by 50% depending on the definition used. Also, sources of registration bias need to be examined, because they differ considerably by country. Underregistration is known to be as high as 20% of perinatal deaths. The next step is to correct perinatal mortality figures according to differences in known risk factors. The perinatal mortality rate then can serve as a reasonable indicator for the quality of antenatal and perinatal care. In western countries, perinatal mortality could be reduced by as much as 25% with improved standards of care. Policies and practices in individual countries concerning ethical issues related to termination of pregnancy and care of newborn infants with (very) poor prognosis need to be taken into account as well. They are not related to quality of care, but do have a relatively large impact on the perinatal mortality rate.