B.C. is here for the record—there is no dispute about it and never has been. B.C. follows the year and follows the word century:
... sometime before 2000 B.C., corn was introduced —Katherine Hinds, Brown Alumni Monthly, October 1982
We have Panini's analysis of Sanskrit from the fourth century B.C. —Edward Finegan, Attitudes Toward English Usage, 1980
A.D. is a different story. It presents three problems: Does it go before or after the year? Can it be used with in? Can it be used after century?
The traditional and still most frequently used styling places A.D. before the year:
A.D. 1942 was the year —Time, 28 Dec. 1942
... objects, which date from A.D. 200 —Newsweek, 10 July 1944
Until A.D. 1200 the Great Plains were virtually unpopulated —Albert H. Johnston, Publishers Weekly, 29 Dec. 1975
Some writers and publishers, however, place A.D. after the date like B.C:
Strictly speaking, we should use A.D. only with numbers indicating particular years (43 A.D., 8-10 A.D,) —MacCracken & Sandison 1917
Lucian flourished approximately 125-190 a.d. — Insect Enemies of Books, 1937
... the vast platform that before 70 A.D. had supported the Temple —John Updike, Bech is Back, 1982
MacCracken & Sandison 1917 finds that usage justifies placing A.D. either before or after the year (they chose after); Reader's Digest 1983 also finds placement after the date acceptable, especially in writing in which such dates are frequent.
MacCracken & Sandison brings up the question of in:
"Though purists insist on 'He died 48 A.D.' [not in 48 A.D.], usage allows in. . . ." The objection to in is based on insistence on the literal translation of the Latin anno Domini "in the year of the Lord." Bremner 1980 is still defending the position of the 1917 purists, but no other commentators mention it.
Insistence on the literal "in the year of the Lord" is also the basis for the objection to using A.D. after century; the use is illogical if you insist on the literal interpretation. Bremner 1980 does. But many people will agree with Johnson 1982 when he terms the etymological objection to A.D. after century "rather a fussy point" and adds "there is little to be gained by binding A.D. forever to its original exact meaning." There is plenty of evidence that writers and publishers have found A.D. convenient to use after century:
... fourth century a.d. —Leonard Bloomfield, Language, 1933
Arabians borrowed coffee from the Abyssinians about the twelfth century A.D. —Science News Letter, 28 July 1945
... came over from Ireland in the second century a.d. —Thomas F. O'Rahilly, Early Irish History and Mythology, 1946
... built in the first half of the third century, A.D. — Current Biography, October 1967
... the first century a.d. —John P. Dessauer, Book Publishing, 2d ed., 1981
Bremner 1980, on the side of the literalists, suggests that only B.C. be used with century; any century without B.C. could then be assumed to be A.D., and there would be no need for the illogical designation. His solution would work were the use of A.D. not already established, but as matters stand it hardly seems a realistic goal. Ebbitt & Ebbitt 1982 report that some sophisticated stylists and Latinists analyze both A.D. and B.C. as non-translated adverbials, applying to both years and centuries; their interpretation sidesteps all controversy.
In summary, B.C. goes after the year and after century; A.D. is more often placed before the year than after; it is widely used after century. Some commentators attempt to rate these stylings on a basis of formality, but our evidence tends to undercut that argument. The question is most likely to be decided as a matter of individual or house style; in other words, consistency of application is more important than which styling is selected.(资料出处：韦伯斯特英语用法词典)