Variations in transgene expression due to position effect and copy number are normalized when analysing and comparing the strengths of different promoters. In such experiments, the promoter to be tested is placed upstream to a reporter gene and a second expression cassette is introduced in a linked fashion in the same transfer DNA (T-DNA). Normalization in the activity of the test promoter is carried out by calculating the ratio of activities of the test and reference promoters. When an appropriate number of independent transgenic events are analysed, normalization facilitates assessment of the relative strengths of the test promoters being compared. In this study, using different modified versions of the Cauliflower Mosaic Virus (CaMV) 35S promoter expressing the reporter gene beta-glucuronidase (gus) (test cassette) linked to a chloramphenicol acetyl transferase (cat) gene under the wild-type 35S promoter (reference cassette) in transgenic tobacco lines, we observed that cat gene expression varied depending upon the strength of the modified 35S promoter expressing the gus gene. The 35S promoter in the reference cassette was found to have been upregulated in cases where the modified 35S promoter was weaker than the wild-type 35S promoter. Many studies have been carried out in different organisms to study the phenomenon of transcriptional interference, which refers to the reduced expression of the downstream promoter by a closely linked upstream promoter. However, we observed a positive interaction wherein the weakened activity of a promoter led to upregulation of a contiguous promoter. These observations suggest that, in situations where the promoters of the test and reference gene share the same transcription factors, the activity of the test promoter can influence the activity of the reference promoter in a way that the test promoter's strength is underestimated when normalized by the reference promoter.