It really depends on what you are trying to accomplish.. AND especially depends on what software you are using. Certain software is just easier to use than others, but have limits.
Since we don't really know what software you are using, or what kind of video/song.. Most software can import video and song, and then it's just a matter of moving the song along the timeline.(to start when you want)
You can also create 'clips' of songs and videos. You will also want to split and merge certain sections of video and create good 'transitives'. (When one scene turns into a different scene.) And do NOT overdo transitions.. that is a major no-no. Just avoid using them repeatedly. (Like every 15-30 seconds is a bad idea.) The ONLY exception to this rule is a photo-video project. The more ideal approach is to create a morph and transition the morphs into the video.. much harder, but the elite.
(Trust me.. I did morph & video my last two years in highschool.. I was the best. Lazy sometimes playing UT instead of working.. but I always had my projects done and 'perfected'.) Some no-nos on morphs are: Flipping the images from left-right or right-left.. big no-no (easily fixed by 'flipping' the photo file instead), using an image with anything near or covering facial features (headphones, etc), using *.jpg files (or other 'lossy' formats), not having 'enough' transition points (more is generally better, but makes the editing more complex), and lastly time.. Take lots and lots of time to work on one.. and more time 'perfecting' it.
But that's just my opinion..
~~Azurith