There are many factors to consider when you're choosing a program. Everyone will have different priorities, but I think there are a few essential points to consider as you make your decisions about where to apply. I've listed them in approximate order of importance to me.
The faculty at your graduate institution will play an enormous role in your academic development. You will need a mentor and a dissertation committee to guide you through the major research project(s) that you're expected to complete. Therefore, it's important to consider potential mentors at every school where you apply. In most cases, you don't have to select your adviser up-front -- life science programs tend to admit students to the program as a whole, then allow several short-term lab rotations before the student chooses a mentor. However, you don't want to apply to a school only to find that you're not very interested in the labs there, or that the dream professor who inspired you to apply is no longer accepting students. For each school that I considered, I spent several days going over faculty websites to determine the major research topics being studied at that institution. Many schools have special research centers composed of many affiliated, collaborating labs. If you're interested in, say, neurodegenerative diseases, you'd do well to apply to a school like Emory, which has a Center for Neurodegenerative Disease. If you're interested in vision research, you will soon notice that many labs at UC Berkeley focus on vision. Of course, plenty of great labs also exist at universities without special research centers dedicated to their work. But as a rule, I didn't apply to any schools with fewer than three faculty that I would consider as possible mentors. This ensured that I would have options even if one of my top faculty retired, stopped taking students, didn't achieve tenure (make note of how many of your favorites are assistant professors, as this can be a concern), or just turned out to be a jerk.
In addition to the research performed in your mentor's lab, you'll be responsible for taking classes, attending seminars, working as a teaching assistant, and passing qualifying exams. Each program tends to vary in the specifics of these requirements, although they often share many similarities. Take a look at your prospective programs' websites and look at the curriculum. For many of the schools where I applied, the first year was dedicated to coursework and lab rotations, with less coursework and more teaching responsibilities in the second year and beyond. One program had no teaching requirement, others required one or two semesters of teaching, some with the option to teach more to earn extra money. The number of required core courses, electives, journal clubs, and seminars can also vary between programs. Additionally, some schools offer special educational resources like grant-writing courses, career development seminars, and special certification programs in addition to your primary degree.
Think about what you want to do when you finish graduate school, and look for programs that will prepare you for that goal. If you want to teach, you might choose a program that provides extensive pedagogical training over one with no TA requirement. If you want to pursue "alternative careers," you can look for programs with certificate options in science writing or public policy. It can also be helpful to look at the electives offered -- if there are a lot of really cool-sounding classes, then this school has faculty with the expertise needed to teach you about all that cool stuff, and makes it a priority to train their students in the subjects that interest you. That said, don't rule a program out just because they require a biochemistry course and you hate biochemistry, or they require a semester of TAing and you don't think you want to teach. Part of your training will always involve exposure to new experiences and techniques, to prevent intellectual stagnation. You may find that you really like something after your program compels you to give it a try.
Finally, consider how your program of interest fits into the graduate school as a whole. In neuroscience, and many other life sciences, some schools isolate their programs, while others combine them into biological/biomedical "umbrella" programs consisting of many different fields related to biology. These large programs can have many positive qualities -- they may allow you to work with mentors not explicitly affiliated with your field, or provide access to more coursework options. However, some people may feel lost in an introductory biosciences class of 100 new PhD students, preferring the intimacy of a small program of 10. My program is affiliated with an umbrella (the Graduate Division of Biological and Biomedical Sciences), but retains some independence. Neuroscience students have some classes, seminars, and events to themselves, while other aspects of the curriculum are shared among all of GDBBS. We recruit and admit our students independently and have some degree of independent program administration. I think this is a good balance, but you may not agree. Above all, what matters is how well the program is able to do the job of educating its students, whether it does this on its own or as part of a larger formal entity.
There are several series of ranking for graduate schools. US News and World Report is perhaps the most well known. The Chronicle of Higher Education also puts out a ranking of PhD programs, although they don't always include everything (there are no 2007 rankings for neuroscience!). You can also check out phds.org to create a customized ranking that weighs schools based on the factors most important to you. These rankings shouldn't be the end-all-be-all in your decision making process. You can still get a good education at a school that is ranked below 50 by US News, for example. However, the rankings may help you to be realistic in your application choices -- you might not wish to apply solely to the schools ranked 1-10, as these will be extremely competitive. Apply to the #1 program if you like, but I suggest checking out schools of middling rank as well. The rankings can help you discover schools that you might not have considered on your own, one of which might turn out to be your dream school.
I mentioned in the Curriculum section that administrators can play a big role in how smoothly a program runs. It can be hard to determine the quality of administration before you visit a program, but there are some clues to be found. Is the website well organized and up to date? Is the application process straightforward? Are people in the program responsive to your questions? A bad website is not reason enough on its own to discount a program, but a good website can sometimes be reason enough to apply! If you get the impression that someone has put a lot of thought into the admissions and recruitment process, the odds are good that they'll be thoughtful about the rest of your graduate education as well.
Because there are so many good graduate programs out there, I think it's perfectly reasonable to limit yourself by geography. Think about the kind of places you've lived, what you've liked and what you've disliked. I grew up in Gainesville, FL and liked it there enough to apply to the University of Florida, but you might not want to live in a sprawling southeastern college town. I went to undergrad at Brandeis University in the suburbs of Boston, and I love the city but was hoping to move somewhere with milder weather. This probably shouldn't be the starting point for your selection process, but if you have a long list of schools that needs to be cut, feel free to cross off the ones in the middle of nowhere, or the ones so far away from your family that you'll never get to visit, or whatever else might be a deal-breaker for you. While the quality of your education is the most important aspect of choosing a graduate school, don't underestimate the qualify of your life outside of school, and the impact it can have on your success. Happy students are healthier, more productive, and better liked by their professors and peers than miserable students.
You may also need to consider schools based on location if you're dealing with a two-body problem. This refers to an academic couple looking for school admissions / job offers in complementary locations. In this case, you'll need to collaborate with your partner and come up with a list of schools together. Feel free to apply to your dream school even if it is far away from any school that your partner would consider, but be aware that you may be faced with a very tough decision if you're accepted. Otherwise, you can concentrate on applying to a lot of the same schools, or different schools that are located in the same city or region. As a former Bostonian, I know that there are dozens of excellent universities within half an hour of each other in that area, which could work out nicely for couples dealing with the two-body problem. I was lucky when I applied to graduate school, since my partner has a job that allows him to find work almost anywhere. I did limit myself to schools near a major metropolitan area, however, to make sure that he'd have plenty of employment opportunities wherever we ended up.