This report based on mortality data from Birmingham from the 1947-56 period analyzed changes over time in the association between infant mortality birth rank and maternal age and in the causes of death (infective noninfective and indeterminate). Analysis of the 1030 postneonatal deaths observed in a series of 128496 births indicated a moderate reduction of mortality with increasing maternal age a more marked increase in mortality with increasing birth rank and a distinct interaction between these 2 factors. Higher mortality was observed among the subsequent infants of women who 1st became pregnant at an early age and/or those whose pregnancies closely folloewed each other. The decline in mortality noted in the period under study was proportionately greater in the groups with the highest incidence of postneonatal deaths. Infective deaths showed a steady although decelerating decline between 1947-56 while noninfective deaths showed little change. It is concluded that improvements in deaths from infectious conditions reflect wide use of antibiotics and other modern treatments; however the decline in mortality appears to be tapering off and future advances will depend on improved understanding of the prevention and treatment of noninfectious diseases.