Little is known about ulcer outcome in the elderly. The aims of the present paper were to establish whether risk factors of slow-healing peptic ulcer can be demonstrated in the elderly and whether clinical differences exist between ulcer patients whose age of onset of the disease was before or after 65 years old. The short-term, open study, involving 1052 elderly patients (over 65 years) in 37 gastroenterology centers throughout Italy aimed to compare two schedules of ranitidine treatment: 150 mg twice daily versus 300 mg at bedtime. As nonsignificant differences were found between these two schedules, the sample was considered as a whole. It included 319 gastric ulcer (GU) patients, 699 duodenal ulcer (DU) patients, and 34 concomitant GU and DU cases. Ninety-three patients dropped out of the trial; 79/294 GU, 138/635 DU, and 10/30 GU+DU were found still unhealed after four weeks and 20 GU, 15 DU, and 1 GU+DU remained so after eight weeks. Statistical analysis was performed using likelihood-ratio and Pearson's chi-squared tests and Cox's models. Univariate analysis showed that the indicators of slow-healing GU were ulcer size (P = 0.002) and persisting ulcer symptoms (P = 0.0001); indicators of slow-healing DU were ulcer size (P = 0.0001), persisting ulcer symptoms (P = 0.0001), alcohol (P = 0.0003), and NSAID (P = 0.0088) consumption. DU patients taking antiplatelet drugs have significantly better results after four weeks and worse results after eight weeks (P = 0.0352).(ABSTRACT TRUNCATED AT 250 WORDS)