This is a very personal issue.

Advantages to working:

• It gives you valuable perspective in grad school.
• Sometimes the company will pay for your schooling.
• Gives you time to consider whether you want grad school at all.
• You can save up some money.

Disadvantages of working:

• You get used to the big fat paycheck, and it's hard to take a 75% pay cut.
• You get out of practice of "going to school".
• If you didn't take GRE as a senior, you forget a lot.
• If you did take GRE as a senior, the scores "expire" after a few years.
• Sometimes you lose "currency" if your job doesn't allow you to learn as you work.

So, if you really know you want grad school, go for it. Otherwise, you might consider working for 2-3 years, and then going to grad school.

• Choose the best programs that you can get into.
• Add 1-2 "safe" schools, that you are pretty sure you'll get into, but you wouldn't hate to be at. Remember that it is possible you'll get into some of your preferred choices, but maybe not with financial support. So, choose good schools, but ones that are perhaps less competitive.
• Pay no attention to the academic reputation of the university. It is essentially irrelevant. What matters is the reputation of the department (program) where you will be applying.
• Do your homework. Get the informational brochures about the department. Find out what faculty are there, and what their research areas are. Look for the ones that are relevant to the kinds of research you are interested in. You need to find a fit between your and their interests. Preferably choose places where there are several possible faculty to work with (what if one of them leaves the department after you get there?).
  • Read their papers. Are you interested in their research?
  • Find Dartmouth alums who are now students there; talk to them or find Dartmouth profs who were grad students there and talk to them.
  • Student/faculty ratio.
  • Success rate.
  • Size of program.
  • What will you learn there?
  • What is required of you there?
  • Who is doing interesting work there?
  • Departmental dynamics...is there collaboration? Departments like this tend to be more congenial.
  • Placement of recent PhDs.
• Look through journals in your area of interest; find out who is doing work in that area and where they are from.
• Talk to Dartmouth professors who are in the same general field as the one you are interested in. Discuss the schools you are considering. Ask them for suggestions. Ask them about particular professors and programs. They have connections, too, which might get you more information.
• Visit the campus. Tour the facilities. Talk to grad students, especially those who work with someone you're considering as a potential adviser. Try to get a feel for the atmosphere, the morale. This is more important than you realize. Consider things such as office space, library support, computing facilities, special research labs or computers, etc. These are important when you try to do research. Try to talk to the relevant professors. Consider the locale, and the cost of living.
• Big, famous schools are not necessarily better. In a big place you might have trouble standing out. Consider being a big fish in a small pond. On the other hand, small places can sometimes be somewhat limiting in terms of resources and variety.

• Many terminal-Master's programs (those where you are not planning to get a PhD, just a Master's) require you to pay tuition and fees. Note that many big companies will pay all this for their employees, sometimes on a part-time basis and sometimes as a year off for school, while still paying you that same huge salary! It's an option. (Same for PhD, too, though a part-time PhD takes forever!)
• On the other hand, in PhD computer science programs, they should pay you to go to grad school. Thus, when you are accepted, they will usually offer you some form of financial package. This usually means that they will pay all your tuition, sometimes the additional fees, and a monthly stipend. The stipend may cover 9 or 12 months of the year, and is usually just barely enough to cover living expenses. You almost never have to apply for financial aid at grad school; they just consider all applicants. Check on each school's policies. These packages come in three basic forms:
  • Teaching assistantship (TA): You have to help run labs, tutor, grade, or even teach. First-year students rarely actually teach.
  • Research assistantship (RA): You help a particular professor with their research, usually supported by a grant.
  • Fellowship: This is the best. Usually this gives you all the money but with no teaching or research opportunities. Some schools do this for many of their first-year students, with the expectation that you will become a TA or RA in later years. A really good deal gives this to you for 3 or 4 or 5 years.
• Definitely plan to apply for third-party fellowships. Many require you to be a US Citizen, but not all. If you can get one of these, they often pay more, usually don't require any teaching or other duties, and you can call up the schools and tell them you have your own funds...which they are definitely psyched about. It also looks great on your resume in the future.
• One thing to know is that your guaranteed student loans (GSLs) are deferred (interest-free) so you don't have to pay them while you're a full-time student. So when you graduate from grad school, and have a big fat paycheck, and inflation has reduced your loans to a fraction of their former value, then you pay them off.

• Most important are the letters of recommendation! The best letters are those that will be enthusiastic, be obvious that the person knows you well, and are from someone the reader can trust (e.g., someone with a reputation in academics, or the department chair, etc.). Choose people who know you well, and who can write a detailed letter. Also try to choose at least one senior prof who has some reputation in the field, or who can otherwise establish their reputation to the letter reader.
  • For most letters, you have a choice of whether to waive your right to see the letter. You should always waive this right. Yes, it may bother you greatly to waive any right, but think about it this way. If you do not waive this right, then your letter writers will not be able to write about you in confidence. Your letter readers will know this, and the letters about you will be discounted. In essence, the one subjective source of information on you has become worthless. So make sure you waive your right to see your letters.
• If you can, include something that makes you special, like a paper you wrote or the description of a programming project or the like. These "appendices" are perfectly fine and will often really make a big difference. Do not send a printout of code. But you might consider pointing to a cool program you wrote, say on github, if you really have something to show for it. Most people won't bother to try it, but they might if you make it easy for them.
• Test scores and grades are somewhat important; bad scores and grades are enough to keep you out, but great scores and grades are not enough to get you in.
• The essay is moderately important; in particular, it had better not be bad. The best essays somehow set you apart. Write your essay carefully, and rewrite it again and again. Tell them what research areas you are interested in, and why. Make it clear that you have a passion for research and would gladly charge through a machine-gun nest for the privilege of doing research. Tune the essay for each place you apply; tell them why you want to go there. Get the essay read over by someone else; make sure it is written really well. Sell yourself! Tell them what makes you unique. If you have written a research paper, or a major computer project, send them the paper (or documentation, if reasonable), perhaps after cleaning it up by making it even better.

• Start early: senior fall, or even junior summer.
• Write to lots of departments and request informational brochures and application materials.
• Take the GREs in October, or at least in December, and have the scores forwarded directly to the schools where you are applying.
• Request transcripts and have them directly forwarded.
• Send in your application well before the deadline.
• Follow up on everything. If you're using snail-mail, be paranoid.  (Who uses snail-mail anymore?) For example, send your application materials return-receipt-requested, and include a self-addressed, prestamped postcard that says "XXX university has received my materials", so that you know when they have arrived. Call them if you do not receive this. Especially: followup on faculty letters of recommendation (be tactful of course), regardless of whether they are submitted electronically or physically.
• Ask for your letters of recommendation early, October if you can. Go ahead and ask even before you have all the forms they will need, or even before you know the complete list of places you will apply. They can start writing the letter and then mail it when you give them the materials. Ask for the letters in person if possible - talk with the prof for a little while. (Make sure they remember who you are! Obviously, you want someone who really knows you well.)