Everybody has there own ways of keeping lab notebooks, but there are some things that should always be present and some notebook-keeping practices which shouldn't be violated. I'm an organic chemist, and this is my version of good practices for keeping a lab notebook.
My method -
1) Before I start to mix chemicals, I have a reaction scheme that shows at a bare minimum the starting materials I will be using and the product I expect. If I have any intermediates that I might be able to detect by TLC or mass spec, it helps to have them in the reaction scheme as well. I also have a table of reactants and products, with spaces for name, volume, density, mass, molecular weight, and equivalents. It helps to have the density of any reagents you will be measuring by volume instead of mass, and I personally like to have the name of the compound - if commercially available, it will be the name on the bottle, and if it is a product from a previous synthesis, it will be the "lot number" of the compound. (Notebook number will be defined later). I also have a reference for the reaction I am running; either a literature source, a notebook page if I'm basing it on my own or my lab-mates notebook reaction, or one I use occasionally, "exp" (for experience), which means I've run the reaction so often before that I'm doing it from memory.
2) The first thing I write down when I start the actual reaction is the date that I am setting it up.
3) The procedure - here there is some difference of opinion. I don't like to write down anything I haven't done yet; if I need that, I will have it on a separate piece of paper. My lab notebook is a record of things that I have done, not things I plan on doing. As I complete each addition or step of a reaction, I will write it done in the procedure.
4) The results - there should be some record of what happened to the reaction, even if all you did was take a TLC and throw the reaction away. Any material I get out of the reaction will get a "lot number" - this is composed of the notebook identifier (since I am in industry, this is assigned by the company when I am issued a new notebook), the page number, and a consecutive number for each product I isolate. For instance, my crude material after workup might be FLED123-45-1, and then when I chromatograph the material and isolate three spots, they will be FLED123-45-2, FLED123-45-3, and FLED123-45-4. If I then crystallize FLED123-45-3, the solid isolated might be FLED123-45-5, while stripping down the mother liquor would give FLED123-45-6. Those numbers will remain attached to that lot of material, and any data collected will have that as a lot number. Any further reactions I run on the products will have that lot number attached to the starting material. The data itself (NMR, MS, HPLC, chromatography traces, CHN, or whatever else I run) is collected in a separate loose-leaf notebook in order of lot number.
5) Once the reaction is complete and nothing further will be written on it, any blank space on the page is marked through and the page is signed and dated at the bottom.
6) Every few days, I find a fellow researcher on a different project to witness and date any pages I have completed since the last time.
THINGS THAT SHOULD ALWAYS BE PRESENT:
1) A reaction scheme. You need to show what materials you started with and what you intended to make.
2) A reference to show where you got the reaction from. This can be very helpful when you have to track down references for publication!
3) A way to trace every chemical you used. If they are commercially available that should be indicated, if they came from a previous reaction, indicate clearly which reaction, and if you borrowed them from a lab mate, ask for a lot number that will let you track it back to their notebook preparation.
4) A record of everything that you actually did in the reaction - the weights you recorded when measuring out your compounds, the temperatures and times that you heated the reaction, how it was worked up and purified, and what data you acquired for each product.
5) An identifier for each of the products of the reaction, so you can follow them in subsequent reactions.
6) If you fill up the page, a note to indicate where the reaction is continued.
7) The dates at which you started and completed the reaction.
DO NOT SKIP PAGES, even if you know the reaction will take up more than one page and you are setting several up at the same time. This makes it very hard to defend your work as a chronological record of research if for any reason the need ever arises. For the same reason, DO NOT LEAVE BLANK SPACE, even if you know how much you'll need for filling in data later. Any space that gets left at the bottom of the page, mark through to indicate that it will not be used at a later date and that the reaction is complete at that point.