In order to detect the abuse of substances in sports, the knowledge of their metabolism is of undisputable importance. As in vivo administration of compounds faces ethical problems and might even not be applicable for nonapproved compounds, cell‐based models might be a versatile tool for biotransformation studies. We coincubated HepG2 cells with metandienone and D3‐epitestosterone for 14 days. Phase I and II metabolites were analyzed by high‐performance liquid chromatography (HPLC)–tandem mass spectrometry and confirmed by gas chromatography–mass spectrometry (GC–MS). The metandienone metabolites formed by HepG2 cells were comparable with those renally excreted by humans. HepG2 cells also generated the two long‐term metabolites 17β‐hydroxymethyl‐17α‐methyl‐18‐nor‐androst‐1,4,13‐trien‐3‐one and 17α‐hydroxymethyl‐17β‐methyl‐18‐nor‐androst‐1,4,13‐trien‐3‐one used in doping analyses, though in an inverse ratio compared with that observed in human urine. In conclusion, we showed that HepG2 cells are suitable as model for the investigation of biotransformation of androgens, especially for the anabolic androgenic steroid metandienone. They further proved to cover phase I and II metabolic pathways, which combined with a prolonged incubation time with metandienone resulted in the generation of its respective long‐term metabolites known from in vivo metabolism. Moreover, we showed the usability of D3‐epitestosterone as internal standard for the incubation. The method used herein appears to be suitable and advantageous compared with other models for the investigation of doping‐relevant compounds, probably enabling the discovery of candidate metabolites for doping analyses.