There is no doubt that some players have used steroids and increased their power because of them, but the fact that the home runs haven't gone away with testing seems to indicate that other factors are probably more responsible for the league-wide home-run surge.
What could have caused the increase in home runs? A likely culprit is expansion. Major League Baseball expanded by two teams in 1993 (Colorado Rockies and Florida Marlins) and two more teams in 1998 (Arizona Diamondbacks and Tampa Bay Devil Rays). Expansion flooded the league with pitchers who were previously not major league quality players, thus providing more opportunities for sluggers such as Bonds to exploit weak pitching. Since the initial expansion, the league has averaged 2.12 home runs per game, compared to 1.58 home runs per game averaged over the preceding 15 years—an increase of 34 percent.
Bradbury, as others before him, also cites smaller parks and changes in equipment. In light of all of this, why steroids dominates the home run conversation while the multi-faceted change in context continues to be ignored is beyond me.