Aims: Currently pulmonary carcinoids are separated into typical and atypical based on mitotic count and presence of necrosis, according to the World Health Organization. At variance with gastroenteropancreatic neuroendocrine tumours, which are graded based on mitotic count and Ki-67 proliferative index, the use of Ki-67 for grading pulmonary carcinoids is still under debate. Methods and results: In this study we evaluated the prognostic impact of Ki-67 assessment in a multicentre cohort of 201 carcinoids [147 typical carcinoids (TCs) and 54 atypical carcinoids (ACs)] using manual analysis (2000 cells counted) and digital image analysis (in-house Leica Qwin program; ≥4500 cells counted). The Ki-67 proliferative index was correlated with overall survival by means of univariate analysis and in comparison to clinical data by means of multivariable analysis. The Ki-67 index was significantly higher in ACs than in TCs for both counting methods (P ≤ 2.7e−5). In addition, using cut-offs of 2.5% and 4% (manual counting) or 1% and 5% (digital analysis), the highest differences in overall survival were observed (P ≤ 0.0067). Nevertheless, histopathological classification into TCs and ACs showed an equally strong association with disease outcome, although Ki-67 had some additive value within TCs. Ki-67 index was not an independent predictor of survival in multivariable analysis. Conclusions: Our study demonstrates that, although Ki-67 is a strong prognostic factor for pulmonary carcinoids, its usefulness in addition to histopathology in prediction of prognosis is limited. None the less, it may have additional value, especially in cases that are difficult to classify, in combination with histopathology and other molecular markers.