Understanding a region's climatic history can help locate areas rich in species, scientists say.
A research group from North and South America used the climatic history of Brazil's Atlantic Forest to pinpoint likely hotspots of genetic diversity.
With frogs, at least, the idea worked, pinpointing places with a rich lineage.
Writing in the journal Science, they say this could set "new priorities" for conservation in regions likely to harbour interesting plants and animals.
"With this method, we can identify areas that have been working as refugia for biodiversity," said research leader Ana Carolina Carnaval from the University of California at Berkeley.
"These are areas that have remained climatically stable through time, where local communities have been able to persist.
"Despite the fact that we haven't sampled them exhaustively yet, we think there is a lot of undocumented, hidden diversity there, the potential for a lot of species still unknown to science."
The Atlantic Forest once stretched for thousands of kilometres down the Brazilian coast, and extended inland through Paraguay into northern Argentina.
Less than 10% of its original area remains - largely fragmented into hilltop groves - and is categorised as a World Biosphere Reserve because of the ecological riches dwelling within.
Dr Carnaval's team used climate models to show that the central part of the forest had seen less climatic variation over the last 20,000 years than the more explored southern region.
This ought to mean, they hypothesised, that species might have survived there undisturbed by climatic fluctuations, whereas in other parts of the forest their existence would have been more transitory.
To confirm the idea, they took DNA from three frog species occurring across the region and found that those in the central part of the forest were more genetically diverse, indicating that populations there had been more stable over the millennia.
If the idea holds true generally, they say, this could help pinpoint areas that researchers could usefully target.
"We think this technique could be applied in other countries and other hotspot areas to identify regions that haven't been well sampled yet, regions that could possibly harbour as yet undiscovered unique diversity," said Dr Carnaval.
"This is a general method for identifying and prioritising hotspots within hotspots, for finding highly diverse areas that have not been fully explored."
The concept of "biodiversity hotspots" was developed by the UK scientist Norman Myers 20 years ago, and is something that conservationists routinely use.
Even so, hotspots can be really big - the Atlantic Forest is an example - so, practically, targeting areas within them as priorities for research or for protection ought to be worthwhile.
"It's really interesting and a good piece of science," said Jonathan Baillie from the Zoological Society of London (ZSL).
"There are all sorts of limitations, but this is definitely a positive thing that gives you one extra view of the situation, and it's a tool that conservation organisations may incorporate into their decision-making."
Some of the limitations probably include the region of the world concerned. Within 20,000 years, for example, many northern lands have experienced widespread glaciation, which wrought much rougher changes on ecosystems.