This study consists of three parts. In the first part the risk of death for widowed persons is studied as a function of time interval since the day of bereavement. The effects of bereavement on mortality are investigated in all widowed people in Sweden (about 360 000) from 1968 to 1978. Among widowers above 65 years of age there are nine deaths per 1000 in excess compared with married men during the first 3 months after bereavement. In comparison with married people in the age group 70-74 it is found that among widows there is an increased mortality by 22% and among widowers by 48% during the first 3 months after bereavement. Further observation, during a period of altogether 11 years, showed that excess mortality continues, though at a lower level. In the second part, causes of death in the age group 70-74, divided according to marital status, are studied on the basis of data from the National Central Bureau of Statistics. The excess in mortality is due mainly to cancer and cardiovascular deaths, but also accidents, suicides and cirrhosis of the liver. The third part deals with differences between marital status groups with respect to tobacco smoking and alcohol abuse by using data from the population study of "70-year-olds in Göteborg" (H 70). By using data from central registers together with data from the population study it is possible to show that life style factors have an impact on the difference in mortality pattern between married and widowed people in Sweden.