Probiotics are viable non-pathogenic micro-organisms which, when ingested, exert a positive influence on host health or physiology. We have critically analysed the evidence for the efficacy of specific probiotic strains in human gastrointestinal diseases. The best evidence can be obtained with randomised controlled trials which avoid bias. Good evidence has been obtained with several strains in the prevention or treatment of antibiotic-associated disorders, in the treatment (and to a lesser extent prevention) of gastroenteritis and acute diarrhoea and in the alleviation of lactose intolerance. We also analysed the recent randomised controlled trials performed in patients with Clostridium difficile or Helicobacter pylori, inflammatory bowel disease, irritable bowel syndrome, non-ulcer dyspepsia and colon cancer.