Young women with type 1 disease were nine times more likely to die than their peers.
As many as 75,000 people with diabetes die each year. Yet almost a third of these could be avoided by better management of the condition, the study found.
Authors of the report said more must be done to effectively manage the condition and reduce loss of life.
DIABETES EDUCATION RESOURCES
Make sure you're up to date with our diabetes CPD and CME resources on myCME.com.
Diabetes UK chief executive Dame Barbara Young described the findings as 'incredibly alarming'. Experts said the rate of deaths would increase unless care improved.
It follows the government's diabetes czar calling on GP practices to deliver more of the nine key annual diabetes checks to improve management of the condition.
Researchers from the National Diabetes Audit linked patient data from the 2007/8 audit to death certificates held by the Office for National Statistics.
This included data from 5,359 GP practices and 1.4 million people with diabetes - around 68% of all those suspected of having the disease in England.
From these data, researchers estimated that mortality among diabetes patients accounted for around 15% of the 460,000 deaths in England each year.
Type 1 diabetes patients were faced with 2.5 times higher risk of dying than the general population when adjusted for age. This compared with 1.5 times higher in type 2 patients.
Taken across all people with suspected diabetes, researchers estimated that as many as 24,000 deaths excess deaths occur each year.
Men fared worse than women, with type 1 males at 30% higher risk and type 2 males at 21% higher risk than their female counterparts.
Excess mortality was far greater for younger adults. Mortality among females with type 1 diabetes aged 15-34 was nine times higher than the background rate for this age group, and four times higher in men.
Mortality rates varied by up to 34% between London and the North West - the highest and lowest rates in England.
The report also found a strong link between deprivation and mortality risk from diabetes.
For type 2 diabetes, those in the least deprived fifth of the population had 1.4 times higher risk compared with the general population, but this rose to 1.8 times among the fifth most deprived people.
Authors said healthcare providers with higher or lower than expected levels of mortality should be investigated.
Diabetes was recorded on the death certificate in just 1.1% of cases. Researchers said death certification practise 'appreciably underestimates' excess mortality caused by diabetes by more than half.
Audit lead clinician Dr Bob Young, consultant diabetologist, said: 'For the first time we have a reliable measure of the huge impact of diabetes on early death. Many of these early deaths could be prevented.
'The rate of new diabetes is increasing every year. So, if there are no changes, the impact of diabetes on national mortality will increase. Doctors, nurses and the NHS working in partnership with people who have diabetes should be able to improve these grim statistics.'
Dame Barbara said: 'These figures are incredibly alarming as there is no reason why people with diabetes cannot live long and healthy lives if they receive the right care and support to help them manage their conditions.
She said self-management was important, but that it was vital that people with diabetes are well supported.
'We know that half of people with type 2 and more than two thirds of people with type 1 diabetes are not receiving the care they need to stay healthy, so it is imperative we take action now to stop even more lives being needlessly cut short.'
The study was conducted by the NHS Information Centre based on the findings of the 2007/8 National Diabetes Audit, organised by the Healthcare Quality Improvement Partnership.