During the past ten years the sale of snuff in Norway has increased, while the sale of smoking tobacco has declined. This has happened parallel to the introduction of smoking restrictions in public areas and growing concern about the health risks associated with smoking. Approximately 300,000 used snuff in 2003/2004. Of these about 140,000 used snuff every day. The majority of snuff users are men under 45 years. The age at starting is decreasing. Female users are still few, but the number is expected to rise. Situations where the consumption rises and the health effects are not clarified, generates new challenges. To systematically review, assess and summarize the epidemiological evidence concerning the health effects and dependency associated with snuff consumption, and the role of snuff in smoking initiation and smoking cessation. Systematic search was undertaken in the Cochrane Library (Cochrane Database of Systematic Reviews, DARE, CENTRAL (earlier Cochrane Controlled Trials Register), HTA (INAHTA)), Medline, Embase, Psycinfo, Cinahl, Swemed, Current Controlled Trials, from 1872 to 7 February 2005. References of identified trials and relevant reviews were also screened. Epidemiological studies were selected if they were systematic reviews, randomised controlled trials, controlled trials, cohort studies or case-control trials, if they included users of snuff or smokeless tobacco, and measured one of the following: nicotine absorption, toxin exposure, mortality, cancer, cardiovascular effects, diabetes, pregnancy outcomes, oral health, dependence, smoking initiation, smoking cessation (or their synonyms). An evaluation was not performed of documentation from animal experiments with snuff or components of snuff. The member of the expert group Erik Dybing disagrees that documentation from animal experiments was not included as a basis for the group’s conclusions. The literature search gave 1040 hits. Two authors independently screened the abstracts, reviewed 207 publications in full text, and assessed the methodological quality of 59 studies using the Norwegian Knowledge Centre for Health Services’ checklists. 29 studies were excluded due to irrelevant problem formulation, outcome or study design. Because the study population, interventions and outcome measures were heterogeneous, a rating system with level of evidence was used. Scandinavian and American studies were assessed separately due to reported differences in some harmful constituents of snuff products sold in Scandinavia and USA. A systematic review was carried out on the 30 studies that met the inclusion criteria. All the studies were of high or medium methodological quality. Many of the studies however 60 included few individuals who exclusively used smokeless tobacco, and combined use of smokeless tobacco and cigarettes appeared frequently. In addition, information about exposure duration and dose was often lacking. There was overall limited epidemiological evidence of health effects associated with smokeless tobacco. There was strong evidence that smokeless tobacco produces dependency. There was conflicting evidence about smokeless tobacco and its role in smoking initiation and smoking cessation. Nicotine is absorbed rapidly into the bloodstream and to the same degree as when smoking. One study indicates that mean carcinogenic nitrosamine uptake into the bloodstream is lower for Scandinavian snuff compared to American snuff, but not in all individuals. Overall mortality is not increased in users of American smokeless tobacco according to one large cohort study. However, one study is not enough to draw conclusions regarding this issue. One study of high methodological quality indicates that American snuff increases the risks of oral cancer among white women. Other American studies are not congruent and have some methodological limitations. Studies of Scandinavian snuff have not found significantly increased risks for oral cancer but the studies lack statistical power to detect a moderately increased risk. Studies of pancreatic cancer have overall included few individuals exclusively using smokeless tobacco, and conclusions cannot be drawn on the basis of these results neither for American nor Scandinavian smokeless tobacco. One Norwegian study including combined users of smokeless tobacco and cigarettes suggest a connection between use of smokeless tobacco and risk for developing pancreatic cancer when controlling for cigarette smoking. There is limited evidence concerning use of American or Scandinavian smokeless tobacco and cancer of the oesophagus, larynx, stomach, lung, kidney, bladder or prostate. One Swedish study among construction workers exclusively using snuff, found an increased risk for myocardial infarction and cardiovascular death. None of the four following studies have found an increased risk for myocardial infarction or stroke among users of snuff in Sweden. A newly published population based study from the north of Sweden did not find an increased risk for diabetes among users of snuff. However, one study is not enough to draw conclusions regarding this issue. A recent Swedish study found an correlation between use of snuff and increased risk for preterm delivery, pre-eclampsia and reduced birth weight. However, one study is not enough to draw conclusions regarding this issue. Use of American or Scandinavian snuff induces oral mucosal changes (snuff dippers’ lesion) located where the snuff is usually placed. The snuff dippers’ lesions are reversible, while other changes in the oral mucosa caused by snuff use (gingival retractions) are irreversible. Portion packed snuff does not give the same oral mucosal changes as loose snuff. Frequent use of either American or Scandinavian snuff will, after cessation, be followed by withdrawal symptoms. Studies analysing American or Scandinavian smokeless tobacco use as a gateway for smoking initiation are not consistent. Conclusions cannot be drawn due to the conflicting evidence. Studies regarding the role of American or Scandinavian snuff in smoking cessation are not congruent. Due to the conflicting evidence conclusions cannot be drawn. One study indicated that use of Scandinavian snuff did not result in higher cessation rates compared to nicotine replacement therapy. However, one study is not enough to draw conclusions regarding this issue.