There's bad news for anyone who thinks vigorous exercise can compensate for a diet that favours junk food over fruit and veg.

Research led by the University of Sydney has found intense exercise doesn't counteract the detrimental effects of a poor diet on a person's mortality risk.

The study found people who had both high levels of physical activity as well as a high-quality diet had the lowest risk of death, showing you can't outrun a poor diet.

The researchers examined the independent and joint effects of diet and exercise with all-cause, cardiovascular disease and cancer mortality using 360,600 British adults from the UK Biobank.

The UK Biobank is a large-scale biomedical cohort study containing in-depth biological, behavioural, and health information from participants.

High-quality diets included at least five portions of fruit and vegetables every day, two portions of fish per week and lower consumption of red meat, particularly processed meat.

For those who had high levels of physical activity as well as a high-quality diet, their mortality risk was reduced by 17 per cent from all causes, 19 per cent from cardiovascular disease and 27 per cent from selected cancers - compared with those with the worst diet who were physically inactive.

Lead author Melody Ding said there was no escaping the conclusion that "both regular physical activity and a healthy diet play an important role in promoting health and longevity".

"Some people may think they could offset the impacts of a poor diet with high levels of exercise or offset the impacts of low physical activity with a high-quality diet, but the data shows that unfortunately this is not the case," Associate Professor Ding said.

A small number of studies have previously found high-intensity exercise may counteract detrimental physiological responses to over-eating.

However, Associate Professor Ding said the latest study "reinforces the importance of both physical activity and diet quality for achieving the greatest reduction in mortality risk".

The study is published in the British Journal of Sports Medicine.