Being able to predict preoperatively the difficulty of a cholecystectomy can increase safety and improve results. However, there is a need to reach a consensus on the definition of a cholecystectomy as “difficult”. The aim of this study is to achieve a national expert consensus on this issue. A two-round Delphi study was performed. Based on the previous literature, history of biliary pathology, preoperative clinical, analytical, and radiological data, and intraoperative findings were selected as variables of interest and rated on a Likert scale. Inter-rater agreement was defined as “unanimous” when 100% of the participants gave an item the same rating on the Likert scale; as “consensus” when ≥80% agreed; as “majority” when the agreement was ≥70%. The delta of change between the two rounds was calculated. After the two rounds, the criteria that reached “consensus” were bile duct injury (96.77%), non-evident anatomy (93.55%), Mirizzi syndrome (93.55%), severe inflammation of Calot's triangle (90.32%), conversion to laparotomy (87.10%), time since last acute cholecystitis (83.87%), scleroatrophic gallbladder (80.65%) and pericholecystic abscess (80.65%). The ability to predict difficulty in cholecystectomy offers important advantages in terms of surgical safety. As a preliminary step, the items that define a surgical procedure as difficult should be established. Standardization of the criteria can provide scores to predict difficulty both preoperatively and intraoperatively, and thus allow the comparison of groups of similar difficulty. •There is not a clear definition of “difficult” cholecystectomy.•National survey to reach consensus on the definition of difficult cholecystectomy.•Predicting the difficulty offers major advantages in terms of surgical safety.