This study aimed at determining the burden of cancer as a cause of death in Basrah over three selected years; namely 1989, 1997 and 2005. All death registries in Basrah city, Districts and sub districts were used as sources of information for data compiling. Every death recorded in the three years was checked for cause of death and all cases for which any type of cancer written as the cause of death were identified. Data related to age, sex, place of residence, type of cancer, place where death was certified and year of death were obtained. In addition, the numbers of total deaths due to all causes in each year were also recorded. The results indicate a total of 297 deaths in 1989, 499 deaths in 1997 and 649 deaths in 2005 could be attributed to cancer. Regarding sex distribution of deaths, slightly more deaths occurred in males (53.8%) than in females (46.2%) with significant rise of cancer in females in 2005. Cancer as a cause of death represents about 5% of all deaths with some degree of increase in 1997 and 2005 as compared to 1989, but no major change in the cancer specific death rates among different years. Geographically, the distribution shows significant but not substantial variation with years. The mean age of dead persons was similar in the three years. The leading cancer deaths were those of lung, urinary bladder, blood, breast, lymphomas and CNS. The overall risk of death is not much different in different years except for a slight increase in 1997 and 2005 in comparison to 1989. Most cancers show stable or slightly fluctuating level of risk of death with time. Slight rise in the risk of death may be noticed in cancers of CNS, blood (leukemia), Bones, lymphomas and Colon-rectum. The researchers suggest that the stable level of mortality could reflect some improvement in treatment based on early diagnosis of many cancers. A study covering at least ten years is highly recommended to establish more sound time trend in cancer mortality.