Background Plant biostimulants constitute a promising environmentally friendly alternative for increasing crop yield and tolerance to unfavorable conditions. Among various types of such formulations, botanical extracts are gaining more recognition as products supporting plant performance. Moreover, novel tools such as cold‑plasma or low‑pres‑ sure microwave plasma discharge are being proposed as techniques that might improve their efficacy. Elucidation of the biostimulant’s mode of action requires complex research at a molecular level. Transcriptional changes occurring after biostimulant spraying might be investigated using RT‑qPCR. However, this technique requires data normalization against stable endogenous controls. Results Here, we tested the expression stability of ten candidate genes in soybean plants exposed to various biostimulants treatment. Selection of the best‑performing reference genes was conducted using four algorithms (geNorm, NormFinder, BestKeeper, and ΔCt method). According to the obtained results, Bic-C2 (RNA‑binding pro‑ tein Bicaudal‑C) and CYP (cyclophilin type peptidyl‑prolyl cis–trans isomerase) showed highest expression stability, while expression of EF1B (elongation factor 1‑beta) fluctuated the most among a tested set of candidate genes. Conclusions Overall, we recommend using Bic-C2 together with CYP for the RT‑qPCR data normalization in soybean biostimulation experiments. To our best knowledge, this is the first comprehensive study of reference genes stability in plants subjected to biostimulant treatment. The results of this study will aid in further biostimulant research in crop plants, facilitating analyses performed on the transcriptional level.