- Read the abstract
- Look at all the figures and read the captions - make some notes to myself about what I expect to see in the paper: What questions do they care about? What tests did they perform?
- Read all of the words in the paper in order, without stopping, without trying to understand anything.
- Read the paper again, this time skipping sections I don't care about, and focusing on the things I find interesting or trying to understand something that popped up in 1 or 2.
This popular article has some similar advice, tailored to non-scientists, and also provides some nice stats on publishing today.