Dengue -- which is in the same family of flaviviruses as Zika -- infects some 390 million people each year in more than 120 countries of the world.
Dengue symptoms are often mild, but more than two million people annually develop dengue hemorrhagic fever -- which can involve severe headaches, pain behind the eyes, rash, pain in the joints, muscles or bones pain, and leaking blood vessels.
More than 25,000 people die of dengue hemorrhagic fever each year.
"And we have to be confident: Dengue is unique and if you don't do it right, you can do more harm than good."
More From This Section
The vaccine candidate, known as TV003, was tested in a group of 48 people -- half of whom received the vaccine, with the other half given a placebo.
TV003 is made by researchers at the US National Institutes of Health (NIH) from a mixture of four weakened but live viruses, targeted to each of the four serotypes of dengue.
Previous research on TV003, which has been in development for 15 years, had shown it worked well at preventing dengue 1, 3 and 4 viruses.
However, the "portion of the vaccine that was designed to prevent dengue 2 did not induce as strong an immune response in people as the other three components," said a statement from Johns Hopkins.
This time, researchers looked beyond antibody response "for the evidence of actual infection: virus in the blood, rash and low white blood cell count."
The 20 in the placebo group all had dengue virus in their blood. Eighty per cent of them developed a rash, and 20 percent showed lower white blood cell counts, suggesting their bodies were fighting an infection.
The study was done in the United States, where dengue does not circulate in the population.