Here are some more suggestions about using theory.
(1) Don't always go for the cartoon version of the theory, the most common use. In other words, don't cite Benedict Anderson just to make the point that nations are "imagined communities." That Derrida "deconstructs Western logocentrism. That Foucault shows that knowledge is power. These are theoretical commonplaces, like the topoi of medieval/renaissance culture. (Carpe diem, or ut pictura poesis.) Instead, look for the theorist's arguments, the surprising things in their thought. Oftentimes the real substance of the theory does not correspond very much to the popular understanding. I've heard some versions of Judith Butler that are so simplistic that they are actually the opposite of Judith Butler. For example, if I hear someone say that Butler proves that gender is socially constructed, then I would say that this person has not understood anything, since that is a position taken in feminism before Butler. What she is saying, among other things, is that not only gender, but sex itself is socially constructed.
(2) As Andrew Shields said, read the theory as you would the literary text itself, with as fine a degree of scrutiny. The theory itself has be to read, interpreted, evaluated. What are its strong and weak points? What are the limits of its applicability?
(3) Make sure you know the theory well enough so that you could explain it in your own words, without using any of the specialized language of that theoretical school. When you write, make sure it's your voice we're hearing, not a mimicked theoretical voice.
(4) What are the nuances of the material that you are working on? How are these nuances preserved / lost in a particular kind of theoretical reading?
(5) Is the theory something you believe in, or a heuristic? Do you have the difference clearly in your mind? How would someone not already sympathetic to your approach respond?