Neoadjuvant treatment has increasingly become an integral part of the multimodal management of patients with pancreatic cancer. In patients who are able to undergo surgery following preoperative therapy, tumour regression grading remains the diagnostic gold standard for the histomorphological assessment of the effect of neoadjuvant treatment. In recent years, however, there has been growing concern about inherent flaws of tumour regression grading systems as well as their imprecise and impractical criteria that result in divergence of practice and lack of interobserver agreement. Furthermore, existing tumour regression systems differ in their defining criteria and thresholds, leading to incomparability of data. In this review, the principles and limitations of the main existing tumour regression systems are discussed, and potential alternative assessment approaches and novel markers are presented.