Obviously you pick one, start with the abstract and finish with the last reference, right?
Nope.
Most papers are written by academics, not professional writers. And they're not written with the idea of getting you hooked within the first paragraph.
Reading a research paper, especially when you've not had much practice at it, can be a tough job. Chances are you could get stuck partway through the materials and methods. So you put it down. A week later, back to those materials and methods. Oh, turgid stuff.
Split it up
Here's a way of splitting up the task to make it better, more useful, and maybe even (hopefully) more interesting, even gripping.
0. Print it out. You're going to read it off the paper so you can write on it, use highlighters, and remember its spatial layout. A lot of your brain is devoted to processing physical space (as opposed to a static screen where one page is identical to the next, where you're not physically holding it, and where there are no coffee stains or any other distinguishing features). Let that brain do its job.
1. First, read the abstract, the intro, and the conclusion. Only. Ask yourself as you're reading: what is this paper about? What did they want to show/measure? Did they do it? You look to see how well those questions are answered in those three sections.
Then take a break. A few minutes, a day, a couple of days.
2. Next time, read the intro (to refresh your memory) and then read the materials and methods section. What did they do? What were the experiments? What pieces of kit did they use?
Take a break.
3. Next you might like to look at the results. What results did they get? How have they analysed their results? How do they rate their results? Don't get bogged down in any statistical methods that they use - you can do that later if it turns out to be important to you.
Take a break.
4. Look at the theory - if that's part of the paper. How much of it do you understand? What would you need to do to be able to understand it all? (Just make a list - and then you don't necessarily have to do it, at least not now.)
Take a break.
5. What did they intend to show and how well did they show it? What do they say themselves in the critical evaluation section? What would they do differently if they were able to do it all again? What did they not show? What do they suggest for further work?
Take a break.
6. By now, you're familiar with the paper. So you can scan it all through in one go if you wish.
7. Put a summary together about the paper. Something maybe like this:
X and Y investigated topic A. They looked at Z phenomenon (covered in theory) in P scenarios (different experimental setups, e.g. elements, substances, techniques used). Their results showed Q (results section), however they didn't have sufficient data points (critical analysis) to show R (shortcomings). It would be interesting to study S (the further work section) in scenario T (what you're going to do for your project).
8. Then add a bibliography entry to your bibliography management software (or use Word).
That's it. On to the next paper.
The whole idea is to have some questions in mind as you're reading - so you're reading to answer those questions, not for entertainment - and to read the various parts of the paper at different times. Sometimes your brain will be into theory/maths, other times you'll just want to read about lasers or spectrometers or hardware or experimental subjects and questionnaires.
Reading a group of papers by theme makes comparisons between them easier too - you're comparing methods, or subjects, or types of analysis, or domains of application.
The papers aren't just pieces of prose, they're actually tools. So don't read them like a novel, make them work for you. Happy reading!