To characterise long-term mortality trends for diabetes in Australia during the 20th century, and to provide suggestions to health policy-makers.A descriptive study was conducted using existing dataset. Deaths due to diabetes, as underlying cause of death, from 1907 to 1998 were tallied, according to the ICD-9. Trends in diabetes mortality (overall population, under-19 and over 40-year-old age groups) by gender were examined.There was a slightly increasing trend in the mortality rate in males over the study period, from 14.38/100,000 in 1907 to 16.05/100,000 in 1998. Among females, it started from 19.6/100,000 in 1907, reached the peak in early 1940s and then decreased to 10.61/100,000 in 1998. There was a reversal sex ratio after late 1960s with mortality rates among males were higher than females after 1969. There was a significant difference in overall mortality between males and females over study period (p<0.001). The mortality trend among the 40 years and over group was similar to the overall population. The death rates for the under-19 group declined significantly over the study period (p<0.001), but no difference between males and females was detected (p>0.05).The application of insulin played an important role in the reduction of diabetes mortality among the under-19-year-old group. New medical treatment methods to reduce chronic complications and other public health interventions could have made a recent contribution to reducing death rates. It is also probable that over the study period there has been an increase in the likelihood of diabetes being correctly diagnosed.