During a one-year-period 1243 medicolegal deaths in the City of Copenhagen were autopsied after the mode of death was provisionally determined by an external medicolegal examination. In 35% of the cases an atuopsy was requested by the police. Without autopsy the mode of death would have been defined erroneously in 10.4% of all cases. This was particularly because of the difficulties in distinguishing between natural and non-natural death. In a rural district of Sealand in the autopsy rate in the same period was 14%. The rate among presumed natural deaths and accidents was particularly low (7% and 26%, respectively) against 23% and 77% in Copenhagen. This means a greater risk for errors in the estimated mode of death compared to Copenhage. This can also mean unequal rights by law. It is stressed that the police, in some rural areas, should be more liberal in requesting autopsies and at least increase them to the Copenhagen level of about 30%. The external medicolegal examination alone offers too little reliability, and it is stressed that such differences from one part of the country to another should not exist in a small well-organised country such as Denmark.