A group of 73 patients suffering from painful alcoholic, chronic pancreatitis, hospitalized from 1971 to 1981, has been analyzed retrospectively. The aim was to assess the effects of alcohol withdrawal and pancreatic surgery on the course of pancreatic pain. The mean number of years during which the patients complained of pain was 3.5 +/- 0.5 (m +/- SEM). At the end of follow-up, 70 p. 100 of the patients did no longer suffer, alcohol withdrawal was obtained in 45 p. 100 and surgery had been performed in 41 p. 100. Continued alcohol abuse did not prevent pain relief: 60 p. 100 of patients continuing to drink at the end of follow-up, did not suffer any longer. One year after pancreatic surgery, pain relief was more frequent, if alcohol abstinence had been obtained before surgery (p less than 0.01). Among the 53 patients followed up to 5 years after the start of pain: a) the cumulative actuarial probability of disappearance of pain was 17 p. 100 at 2 years, 52 p. 100 at 5 years, 62 p. 100 at 8 years after the start of pain. Alcohol abstinence and surgery were observed during the first five years of pain; b) the mean number of years of pain was lower among the patients who became abstinent early (less than 4 years after the beginning of pain) than among those who did not (p less than 0.01).(ABSTRACT TRUNCATED AT 250 WORDS)