A decade ago I picked up a copy of the La Berge book that Ferris mentions. I highly recommend it as it seems to be the canonical book on the topic since La Berge was credited with being first to medically prove in a lab that lucid dreaming even exists. It takes like 3 months of daily dedicated practice to get the knack, but once you've trained your mind to do it, it gets easier to induce with repetition. Regular lucid dreaming is an incredible experience, and I can't even begin to explain the impact it has had on my own creativity and self of self. If you haven't done it--GO DO IT!!

And I agree with what others said about using LD'ing to aid in learning. For me it is hugely useful to get unconstrained "practice" time to quickly go through steps I need to remember when awake. I call it dream kata. I mean both mental steps, like learning math proofs in my case, and steps for physical things, like learning to rock climb(falling off cliffs and flying away is seriously fun). What's weird is after routinely LD'ing such practice scenarios, when I go back to regular dreaming, I will have non-lucid dreams of doing the same kinds of practices. We dream about what we regularly do each day, so it's a way of tricking the mind to watch and mimic itself. Somehow it engages a part of the subconscious to constantly practice and learn.

While LD'ing sounds trivial, the time can add up and I think it has huge potential to give those who employ it an edge. Let's say you lucid dream for 2 hours every night for just 3 years--then you've got 2k+ hours of practice time in whatever you do. Take Gladwell's book Outliers and his account that "geniuses" in any field excel simply because they have had so much more raw practice than others. While we don't know how common lucid dreaming is amongst these "geniuses", it might be another explantion of this "practice == genius" phenomenon. The way I see it, if you can get 20% of the way to 10,000 hours just in your sleep, there's no reason to not be doing it.