It is important for children to take regular exercise
Using a child's body mass index (BMI) as a measure of the success of exercise targets may be misleading, say experts.
UK researchers could find no difference in BMI between those exercising regularly and those missing targets.
Writing in the Archives of Disease in Childhood, they said blood testing might be the only way to measure exercise benefits.
However, another scientist said BMI did offer useful information, and testing should continue.
The government recommends that children have an hour's moderate exercise or more every day to reduce the chance that they will become obese adults, with a higher risk of diseases such as diabetes or heart disease.
BMI is used successfully in adults as a guide to overall fitness and the success of diet and exercise programmes, but there has been debate over its effectiveness when modified for use in children.
Some experts have suggested that it is perfectly possible for an individual child to be "fat and fit", provided they are sufficiently active.
The latest study, carried out at Peninsula Medical School in Plymouth, examined the exercise levels of 113 boys and 99 girls, born in 1995 and 1996, over a four-year period.
All were fitted with devices called accelerometers, which measure every movement to give an accurate picture of the amount of exercise they did.
In common with other studies, they found that just over half of the boys, and nearly nine out of 10 girls, fell short of the "hour a day" target.
However, despite the variety of different exercise patterns, there was no impact on the relative BMI of the children.
Even techniques used successfully in adults to predict health by measuring fat levels failed to differentiate between those meeting the target and those doing far less activity.
This, however, did not mean that the children meeting the target were no more healthy - blood testing revealed clear differences in the underlying metabolic signs of health - such as insulin resistance and cholesterol levels.
Diet not exercise
Professor Terry Wilkin, who led the study, said: "BMI just doesn't pick up any differences in children - it's just not a sufficiently sensitive measure.
"And you can't carry out blood testing on this scale in schools."
He said other research suggested that it was almost impossible to change the overall amount of exercise taken by any child over a week, even if large amounts of sports were arranged at schools.
"It could be that changing the diet of children will be the modifiable factor."
Dr Richard Winsley, a researcher in child exercise science, said that in the absence of another practical test, BMI should continue to be used, as it still revealed a useful overview of the improving or declining fitness of large groups of children.
He said: "Whether or not BMI is very good for an individual person may be questionable.
"But until someone comes up with a more successful, and equally practical alternative, then it is the best we have got."
He said that physical activity in children could only benefit them, so parents and schools should continue to encourage it.
"A lack of evidence doesn't mean that something doesn't actually work."