Genetically modified primates that glow green and pass the trait on to their offspring could aid the fight against human disease.
Though primates that make a glowing protein have been created before, these are the first to keep the change in their bloodlines.
Future modifications could lead to treatments for a range of diseases.
The "transgenic" marmosets, created by a Japanese team, have been described in the journal Nature.
The work raises a number of ethical questions about deliberately exposing a bloodline of animals to such diseases.
Scientists have managed to modify the genes of many living organisms in recent years, ranging from bacteria to mice.
Mice have been particularly useful experimental models for studying a wide range of human diseases as modified genes are passed on from parents to progeny.
However, mice are not useful for some human diseases because they are not sufficiently similar to produce effects that are meaningful to human disease. Studies of mice with Alzheimer's disease, for example, were stymied simply because their brains were too small to scan at sufficient resolution.
Now, Erika Sasaki of the Central Institute for Experimental Animals in Japan, and her colleagues, have introduced a gene into marmoset embryos that allows them to build green fluorescent protein (GFP) in their tissues.
The protein is so-called because it glows green in a process known as fluorescence.
GFP was originally isolated from the jellyfish Aequorea victoria, which glows green when exposed to blue light.
The protein has become a standard in biology and genetic engineering, and its discovery even warranted a Nobel prize.
From 91 embryos, a total of five GFP-enabled transgenic marmosets were born, including twins Kei and Kou ("keikou" is Japanese for "fluorescence").
Crucially, the team was able to show that their method is maintained in the family - or germline.
They used the sperm from a member of the first generation of transgenic marmosets to fertilise an egg from a normal animal. A significant proportion of the resulting offspring also glowed under ultraviolet light.
Until now, efforts to establish transgenic lines of primates have been unsuccessful. In 2001, a team at the Oregon Regional Primate Research Center, US, succeeded in creating a rhesus macaque that expressed GFP.
Last year, a team at Yerkes National Primate Research Center, Atlanta, US, created rhesus macaque monkeys with Huntington's disease. Four of those are still awaiting puberty, and the researchers hope that they will produce a second generation of macaques with the disease.
The new method improves on previous work using so-called "retroviruses".
These virus "vectors" were added to a soup of sugary solution and this was in turn injected into the monkey embryos.
Although the work demonstrates the principle that a gene can be introduced into a primate bloodline, study co-author Hideyuki Okano of the Keio University School of Medicine said it may not be suitable for studying all diseases.
"We can just introduce genes by virus vectors," he told BBC News, "so the limitation comes from the sizes of genes that can be carried by the retroviruses."
That limitation is about 10,000 bases, or letters, of the genetic code. That upper bound will constrain the diseases that can be studied.
Genes implicated in Parkinson's disease and amyotrophic lateral sclerosis (ALS, a form of motor neurone disease) may well be suitable.
However, genetic regions implicated in Huntington's disease might not fit into a retrovirus.
The work has raised a number of ethical questions about the use of primates in disease research.
Marmosets are New World monkeys and therefore more distantly related to humans than, for example, chimpanzees. But they are particularly useful for the study of disease because they reproduce often and from a young age.
Jarrod Bailey, science consultant to the British Union for the Abolition of Vivisection (BUAV), says he is "disappointed" both ethically and scientifically with the results.
"This sort of research on animals as sentient as monkeys and apes does not have public support," he told BBC News.
Furthermore, he thinks the underlying science is flawed. Animal researchers, he said, "have failed in research to find treatments for Aids, for hepatitis, for malaria, for strokes. Many treatments for strokes work in monkeys but don't work in humans."
"Monkeys do not predict human response and do not tell us about human disease," he added.
However, scientists argue that, because primates are more similar to humans than mice, they present a more refined model of human disease. This would allow scientists to test treatments more effectively, meaning that fewer animals need be experimented on in the long run.
"This experiment is reminiscent of the exciting early days of transgenic research where it was initially difficult to fully know what the potential applications and future research directions might be," said Mark Hill, a cell biologist at the University of New South Wales in Australia.
"As always in this area of research, there needs to be a close linkage between the scientific work, ethical issues and regulatory guidelines."