The American Cancer Society has launched the largest cancer study in American history with more than 1 million persons participating to determine the relationships between environmental factors, life style and cancer.

The study, which will take at least six years to complete, is similar to an earlier cancer prevention study undertaken by the society from 1959 to 1972. That study helped produce the conclusive link between cigarette smoking and cancer.

In the new study, called Cancer Prevention Study II, more than 1 million adults--including 50,000 from Maryland, 30,000 from Virginia and 3,000 from the District of Columbia--will be asked to fill out a four-page questionnaire similar in certain respects to the family history physicians take from new patients.

In addition to questions about the individual's personal health and that of his or her family, questions are asked about smoking, diet, medication and vitamins, occupations and miscellaneous subjects.

The questions about smoking are not limited to smokers, but also ask nonsmokers how many hours a day they are exposed to the cigarette smoke of others, and where (home, work, or elsewhere). Smokers are asked detailed questions about the frequency and type of cigarette they smoke.

The object of the study, according to Lawrence Garfinkel, the society's vice president for epidemiology and statistics and director of cancer prevention, is to determine what factors in the environment and contemporary lifestyles contribute to cancer.

The study's unusually large size, according to Garfinkel, will allow reliable conclusions to be drawn from cross-indexation of various categories of data gathered. Researchers normally consider it sufficient to have a randomly selected survey population of 1,500 to establish a representative sample of the American public. A sample of that size, however, presents problems when efforts are made to draw conclusions about subgroups within it.

The new study is being undertaken, at a cost of about $12.5 million, Garfinkel said, because of new information about possible causes of cancer. The new study asks about consumption of caffeine and saccharin, use of oral and other contraceptive devices, use of hair dyes, mouthwash, and exposure to a variety of substances including asbestos, formaldehyde, hydrocarbon fuel exhaust, textiles, X-rays and other substances.

The study also will attempt to determine whether persons who have close social contacts with friends and family have lower mortality rates than "loners."

In addition to looking at cancer, the study will, like the earlier study, focus on heart disease and stroke, according to Garfinkel.

Subjects for the study are being selected by American Cancer Society volunteers from adults 30 or older. The volunteers, Garfinkel said, have been asked to ask close friends and relatives to participate so that the volunteers are more likely to know where the subjects are every other year for the next six years.

The society will contact the volunteers in two, four and six years to ask them whether the subjects are alive or dead. If the subjects are dead, the society will gather information from the person's official death record and other sources about the circumstances surrounding the person's death.

The survey was begun Sept. 1 and will be updated in 1984, 1986 and completed in 1988 unless it is extended.