Abstract Background The aim of this study was to compare the results of conventional open haemorrhoidectomy as currently practised in Italy (group 1) with stapled haemorrhoidectomy using a 33-mm circular stapling device (group 2). Methods One hundred patients with symptomatic third- and fourth-degree haemorrhoids were enrolled by five hospitals. Patients were allocated to the two groups according to a centralized randomization scheme featuring five permutated blocks of 20. Preoperative clinical examination and anorectal manometry demonstrated no features of anal incontinence. Patients had a clinical and manometric re-evaluation after operation and were asked to complete a clinical diary. After a median of 16 (range 8–19) months patients were administered a standardized questionnaire by telephone. Results Postoperative bleeding requiring haemostatic procedures occurred in three patients in each group. Patients in group 1 complained of moderate pain for a median of 5·3 (range 0–19) days compared with 3·1 (range 0–10) days in group 2 (P = 0·01), while severe pain was present for 2·3 (range 0–24) days in group 1 but only for 1 (range 0–14) day in group 2 (P = 0·03). The median hospital stay was 2 days in group 1 compared with 1 day in group 2 (P = 0·01). In the early days after operation, patients in group 2 had greater difficulty in maintaining normal continence to liquid stools (P = 0·01), but after 30 days the continence score was better in group 2 (P = 0·04). Conclusion Stapled haemorrhoidectomy is as effective as conventional haemorrhoidectomy. Reduced postoperative pain, shorter hospital stay and a trend toward earlier return to work suggest short-term advantages for the stapled technique.