Here is a cautionary tale to illustrate the warnings just offered about removing the guard without paying attention to the side consequences. Ask what White threatens; see that his bishop on b2 attacks a Black knight and that his rook on d1 attacks a Black bishop. The Black knight guards the Black bishop. The position seems to call for a classic removal of the guard: White exchanges minor pieces with 1. BxN, BxB; and then comes RxB, with White winning the loosened bishop on d7.
But when you imagine this or any sequence of captures, pause to consider how the board will look when the action is over. What will be loose? What lines will be left open? There may be a kicker you can play at the end—or a kicker for your opponent. In this case notice that after the first pair of moves Black is left with a bishop aimed at White’s rook. So after White plays RxB, Black plays BxR. Would White then be able to recapture in the corner with RxB? No, because his rook on d1 would have been used to take Black’s bishop. So the attempt to win a piece by capturing the guard would end up losing the exchange.
This position is worth a good look, as it shows the importance of carefully visualizing each piece vacated from its old square and on a new one. It also shows the importance of considering all of your opponent's possible recaptures; a rudimentary way to blunder here is to imagine that after 1. BxN, Black would reply g7xB. BxB obviously is stronger, since it not only takes aim at White's rook but more generally makes the bishop active and keeps the pawn cover in front of Black's king intact.