Round-robin testing for two samples, Reference Sound Source in accordance with ISO 6926 and Office printer with electrophotographic engine, were executed by seven testing laboratories in Japan on 2020. All tests were executed with parallelepiped measurement surface in hemi anechoic chamber in accordance with testing standard for engineering-grade sound power determination, ISO 3744:2010. The results show that sample standard deviation for RSS is better than printer. Standard deviations for overall A-weighted sound power level for two samples are better than combined standard uncertainties calculated with reference example of standard deviations in ISO 3744 (1.5 dB for Reference Sound Source as standard deviation of operating/mounting condition is negligibly small, 1.6 dB for printer as stable operating/mounting condition 0.5 dB). This paper also indicates tips for those who would conduct round-robin testing to obtain valid results by obviating incorrect operations and malfunctions of printers or similar equipment from the experience of some round-robin tests.