Also, work from the back of the synthesis, it's often easier.
Another thought is to remember the fact that most synthesis problems can get a lot of partial credit. If you know what a part of the synthesis ought to be then put it down. If you can't remember how to get from the beginning to where you know stuff or from where you know stuff to the end, it's okay to just put a blank arrow - at least you will get credit for what you know.
Third piece of advice. Don't stare at the problem too long, start writing, it's a lot faster to see if you are on the right track if you try and draw out your idea than if you just try to think through it. Nothing wrong with crossing things out and starting over.
Edit: 4th advice: look through the rest of your exam for ideas about what to use in your synthesis. Sometimes you know a reaction well and solved a predict the products or whatever problem early on but then you completely forget about that reaction when your get to the synthesis part.