Abstract
A method of carrying out a Perinatal Mortality Survey in a non-academic district obstetric service is described. The results show a trend of improving perinatal mortality and the various factors that may be influencing this trend are discussed.