• Check to see if a deadline is on a Friday, this happens all the time and it’s probably a bad idea because people make plans for the weekend and if something should go wrong you’re in for a challenge. I like deadlines on Mondays. As David points out, midweek is probably a lot better, that lowers the risk of having to work in weekends even more.
• Are you responsible for another project that has important events at the same time? Act proactively and early if this happens, reschedule or ask a colleague to help out.
• Take a look at the vacation planner, just maybe that key stakeholder or that rockstar server administrator the project depends on will be off rock climbing in the Alps that week. Ask HR.
• Is the deadline on an official holiday? For other countries (when working with offshore companies) you can probably find an official calendar at iCalShare (compatible with a lot more than just iCal)
• Does the delivery process take a long time? Think about this, some deployments or integrations take a long time to prepare or execute.
• Make sure you have the correct acceptance criteria in your hands. Call you client some time up front and tell him what the status of the project will be at the moment of the deadline, even if it’s already in the project charter. If something is still ambiguous at a late moment it should be defined better, right away.

A classic example of last minute ambiguity is the “time of day” discussion, aka “we meant ready by morning on the day of the deadline, not 6pm because our VP needs to sign off the campaign for the final go live”.

Remember, you can ask all vital planning information you need from any stakeholder once you own the project.