Odd-shaped rocks in the Pilbara region of Western Australia offer compelling evidence they were built by microbes 3.43 billion years ago, scientists say.
The Pilbara structures were first described almost 30 years ago
The structures, known as stromatolites, could only have taken the forms they have if bacteria had been present, a Sydney-led team tells Nature journal.
The rocks' origin is disputed, with some claiming purely chemical processes could have made them.
But the Nature study suggests the biological explanation is the simplest.
"For all these shapes to be formed a-biologically would have required highly unusual and unexpected chemical processes to occur simultaneously in this [one location]," said Abigail Allwood from the Australian Centre for Astrobiology.
"It just becomes ridiculous to support that hypothesis; especially when the biological explanation is so readily acceptable."
Ms Allwood and colleagues have made an extensive survey of a 10km (six miles) stretch of land not far from the town of Marble Bar.
The area is now well inland but shows clear evidence of having been covered by a shallow sea in the ancient past.
The researchers have detailed an array of unusual sedimentary structures - seven clear types in all. Some look like upside-down ice cream cones; others resemble egg cartons.
These laminated structures have been described as stromatolites - the rock piles that in more recent settings are known to have been built by mats of microbes capturing grains and sticking them together.
But the Pilbara structures, found 30 years ago in a rock formation called the Strelley Pool Chert, are controversial.
Claims for individual microfossils of the original organisms within Pilbara's stromatolites have been challenged; and some scientists prefer an entirely non-biogenic explanation for the structures' creation.
These dissenters believe the piles resulted from the chemical precipitations that occurred around undersea volcanic vents.
Allwood's response has been to describe the complexity of shapes and explain how these forms can be linked to different environmental niches in a shallow-sea reef setting.
"We have found an ecosystem-scale remnant of the early biosphere. It's not just a couple of individual or isolated fossils or dubious structures; it is in an entire, pretty well intact, section of hundreds of thousands of stromatolites in a reef ecosystem," Ms Allwood told the BBC News website.
"With that we now gain insight into the conditions that nurtured early life - the biological responses to different environmental processes."
The Pilbara stromatolites are not the oldest claim for life on Earth.
The variety points to an entire ecosystem, say the scientists
Some researchers argue that rocks at Isua in Greenland show the imprint of life at least 3.75 billion years ago.
At that time, these rocks were also on the sea bed. Thin layers of black sediment, separated by distinct layers of volcanic ash, look like they could be composed of the debris of ocean-dwelling microbes.
There are no fossil forms, but the nature of the carbon is consistent with the idea it was processed by living organisms. There are no known older remnants of the Earth's surface than the Greenland rocks - which probably makes Isua the closest science can ever get to the first life.
Researchers are keen to trace the story of the first microbes on Earth because it should provide clues in the hunt for possible life elsewhere in the Solar System.
The type of study conducted on the Pilbara stromatolites might, for example, help scientists interpret similar structures on Mars, should rovers sent to the planet ever come across them.
Mars rovers should look out for stromatolite-like structures
Commenting on this dimension, Dr Ian Crawford, a planetary scientist at Birkbeck College, London, UK, said: "Searching for ancient stromatolite-like structures such as those reported by Allwood et al should certainly be high on the list of future exploration strategies.
"However, given the amount of fieldwork performed by Allwood and colleagues, it must be doubtful whether purely robotic exploration of Mars would be able compellingly to identify such features in the field, and in the longer term effective Martian palaeontology may necessitate human exploration of the planet."