Ban Xiengdi. Site 3 was referred to by Colani as ‘Ban Soua’ and is located a few kilometres to the southwest of Site 2 and consists of seven separate jar clusters. These jar groups range in size from a single jar, to more than 100 jars in the primary group. The groups at Jar Site 3 are characterised by a higher proportion of diagnostic variability, and a larger ratio of stone discs, when compared to Jar Site 1 and Jar Site 2.
According to Van Den Bergh et al.'s account the jar site groups are located on lower hill slope spurs and overlook the expansive plain. The sandstone quarry of the jar resources has been located further up the hill slopes (Site 8).
The groups are located within 6-700 metres of each other (with the exception of Group VI which lies at a distance of 1,200 m).
The site was resurveyed by a PJARP team led by Andrew Ball.