Scientists in Japan and Australia have developed a blood test that can detect the build-up of toxic proteins linked to Alzheimer's disease, reports BBC.
The work, published in the journal Nature, is an important step towards a blood test for dementia.
The test was 90% accurate when trialled on healthy people, those with memory loss and Alzheimer's patients.
Experts said the approach was at an early stage and needed further testing, but was still very promising.
Alzheimer's disease starts years before patients have any symptoms of memory loss.
The key to treating the dementia will be getting in early before the permanent loss of brain cells.
This is why there is a huge amount of research into tests for Alzheimer's.
One method is to look for a toxic protein - called amyloid beta - that builds up in the brain during the disease.
It can be detected with brain scans, but these are expensive and impractical.
The new approach, a collaboration among universities in Japan and Australia, looks for fragments of amyloid that end up in the blood stream.
By assessing the ratios of types of amyloid fragment, the researchers could accurately predict levels of amyloid beta in the brain.
Significantly, the study shows it is possible to look in the blood to see what is happening in the brain.
Dr Abdul Hye, from King's College London, said: "This study has major implications as it is the first time a group has shown a strong association of blood plasma amyloid with brain and cerebrospinal fluid."
The test is cheaper than brain scanning, "potentially enabling broader clinical access and efficient population screening", according to the study.
At the moment there is no treatment to change the course of Alzheimer's, so any test would have limited use for patients.
However, it could be useful in clinical trials.
Prof Tara Spires-Jones, from the Centre for Discovery Brain Sciences, at the University of Edinburgh, said: "These data are very promising and may be incredibly useful in the future, in particular for choosing which people are suited for clinical trials and for measuring whether amyloid levels are changed by treatments in trials."
Dr Hye added: "Considering Alzheimer's disease has a very long pre-clinical phase, a truer test will be how well this test performs in independent, healthy, cognitively normal individuals or even in individuals in the early stages of the disease.".