AbstractIn 2017, various orange coloured petunia on the market turned out to be genetically modified (GM) without an official authorization for commercialization. Sequence analysis suggested these undeclared plants most probably originated from a plant transformation experiment performed in the 1980s. For a deeper understanding how GM petunia entered classical breeding programmes worldwide, and whether they originated from a single source or not, we undertook a molecular genetic characterization of the T-DNA integration sites in different GM petunia cultivars and breeding lines. By means of genome walking, we isolated different T-DNA sequences, which are located at the junctions between the T-DNA(s) and the petunia DNA. Based on the results obtained we conclude that there are at least two T-DNA copies of different lengths. This is supported by Southern blot analysis. For T-DNA1, the 3′-junction sequence was isolated, whereas the 5′-junction remained unclear. In contrast, for T-DNA2, the 5′-junction sequence was isolated, whereas the sequence isolated from the 3′-region consists only of T-DNA, but did not include the junction from the T-DNA to the petunia DNA. We developed primers for event-specific PCRs and screened a set of three orange GM petunia cultivars and 126 GM offspring from a commercial breeding program. We show that both T-DNA copies are present in all our tested GM petunia samples, which underpins the assumption of a single transgenic origin of the undeclared GM petunia. Most likely, the two T-DNAs are integrated in close proximity into the petunia genome.