Background: Data on cancer risk in patients with long-segment Barrett’s oesophagus (BO) from older studies are often difficult to interpret, since the definition of BO has evolved from an endoscopical to a histological diagnosis. In this work the diagnoses in the Rotterdam BO cohort on current standards was redefined to obtain more accurate data on cancer risk in patients who had not undergone standard endoscopic surveillance. In addition, it was determined which patient factors present at index endoscopy were associated with neoplastic progression in BO. Methods: The Rotterdam BO cohort comprises all patients with ≥3 cm BO, diagnosed at endoscopy between 1973 and 1984. In the present study, only patients with intestinal metaplasia were included (n = 105). Follow-up data were obtained by questionnaires and/or interviews with patients or treating physicians. A Kaplan-Meier analysis was used to estimate 20-year risks. Results: The mean length of the BO was 7.1 cm (range: 3-15 cm). Cancer in BO developed in 6/105 (6%) patients, and high-grade dysplasia (HGD) in 5/105 (5%) patients during 1329 patient-years of follow-up, which equals one cancer case per 221 patient-years and one HGD case per 266 patient-years. After a mean follow-up of 12.7 years, 72 (69%) patients had died; only 4 of them died of oesophageal cancer or its treatment. A longer length of BO was associated with an increased risk of progression to HGD or cancer (P < 0.02). Six of 24 patients who ever had low-grade dysplasia progressed to HGD or cancer 2-16 years after a diagnosis of BO. Conclusions: The annual risk of developing HGD or adenocarcinoma in patients with long-segment BO is 0.83%. Death due to adenocarcinoma is, however, uncommon, even in a cohort of patients with longsegment BO. Background: Data on cancer risk in patients with long-segment Barrett’s oesophagus (BO) from older studies are often difficult to interpret, since the definition of BO has evolved from an endoscopical to a histological diagnosis. In this work the diagnoses in the Rotterdam BO cohort on current standards was redefined to obtain more accurate data on cancer risk in patients who had not undergone standard endoscopic surveillance. In addition, it was determined which patient factors present at index endoscopy were associated with neoplastic progression in BO. Methods: The Rotterdam BO The cohort was included in all patients with ≥ 3 cm BO, diagnosed at endoscopy between 1973 and 1984. In the present study, only patients with intestinal metaplasia were included (n = 105). Follow-up data were obtained by questionnaires and / or interviews with patients or treating physicians. A Kaplan-Meier analysis was used to estimate 20-year risks. Results: The mean length of the BO was 7.1 cm (range: 3-15 cm). Ca ncer in BO developed in 6/105 (6%) patients, and high-grade dysplasia (HGD) in 5/105 (5%) patients during 1329 patient- years of follow-up, which equals one cancer case per 221 patient- years and one HGD case per 266 patient-years. After a mean follow-up of 12.7 years, 72 (69%) patients had died; only 4 of them died of oesophageal cancer or its treatment. A longer length of BO was associated with an increased risk of progression to HGD or cancer (P <0.02). Six of 24 patients who ever had low-grade dysplasia progressed to HGD or cancer 2-16 years after a diagnosis of BO. Conclusions: The annual risk of developing HGD or adenocarcinoma in patients with long-segment BO was 0.83%. Death due to adenocarcinoma is, however, uncommon, even in a cohort of patients with longsegment BO.