The present estimate of global mortality caused by cardiovascular diseases is accompanied by a considerable degree of uncertainty, which, in so far as monitoring their emergence in developing countries is concerned, undoubtedly represents one of the major obstacles to effective public health interventions for their control. In much of the developing world, vital registration data are lacking and it would be unreasonable to expect rapid progress in the recording of causes of death because resources are so limited. The most promising avenue is that of the progressive implementation of clearly defined mortality surveillance systems that cover all deaths and permit the attribution of probable causes via lay reporting. The reliability of the data largely depends on the specificity and clarity of the verbal autopsy algorithm employed and on the availability of medically trained personnel to validate the returns.