The Titri fault is about 50km long and runs northeast from the Milton area to the southwestern outskirts of Dunedin City.
At its closest, it comes within about 10km of the city centre.
Movement on the Titri fault is vertical rather than horizontal, with upthrow on the southeastern side of the fault.
Project leader David Barrell, of GNS Science in Dunedin, said the excavation work, a first for the Titri fault, was ‘‘definitely a buzz'', and was starting to confirm some of researchers' hunches.
Previous investigations indicated this fault was potentially capable of producing an earthquake of up to magnitude 7.2.
Full results from the latest research would not be available until next year, but the excavation suggested movement on the fault had taken place, probably more than 40,000 years ago, and further movement perhaps only a few thousand years more recently.
Before the latest research began, geologists did not have enough information to know with confidence when the fault last ruptured, and if the ruptures were evenly spaced, or could occur in clusters.
Three organisations - GNS Science, the University of Otago and Victoria University of Wellington - have together gained about $70,000 in funding from the Earthquake Commission's biennial grants programme to investigate the fault.
They dug a 4m-deep trench earlier this month across the fault in two places: near Milton and the other near Clarendon, north of Milburn, to identify the size and age of its most recent movements.
The two most recent ruptures involved compressional (thrust) movement, totalling about 4m of vertical offset, pushing the ground up on the southeastern side of the fault.
Mr Barrell said the Titri fault was not included in the national seismic hazard model, used to calculate New Zealand's earthquake risks.
It was likely the latest research could eventually result in this fault being included in that model.
However, because initial results suggested the rate of activity on the fault was relatively low, any increase to Dunedin's earthquake risk would probably be ‘‘very slight''.
Near-coastal areas of the southeastern South Island, including Dunedin, were regarded as being, on average, one of the lowest risk areas of New Zealand for strong earthquake shaking, he said.
Mr Barrell said quantifying the fault's behaviour and that of other active faults in the area, including the Akatore fault, would lead to a more accurate estimate of earthquake hazard for Otago.
It would have positive spinoffs, including better guidance for structural engineering.
