Speaking to the faculty in my program, I've noticed that they seem to evaluate themselves and others using two criterion: (1) publication quality and quantity, and (2) ability to obtain grant money. No one discusses grades or performance in graduate school classes. Even when potential faculty candidates come around, this doesn't seem to be emphasized at all. What role do grades play in my PhD career?
As far as your research stature is concerned, grades would matter least of all, below other non-academic stuff as your soft-skills, your personality etc. There are a number of reasons for this:
No one cares about your GPA once you are a researcher! While it certainly looks nice to have a stellar GPA, it's the work that you do and where you publish that would matter. Look up some resumes of notable faculty in your field - how many even list their MS/BS grades?
I might even say that your adviser would not be too happy if you have a 4.0 GPA - as it means you are spending time on perfecting your grades which is more profitably spent on research! I actually have read this on a faculty/university webpage - would post the link once I dig it up. EDIT: Haven't found the faculty website link yet, but here's a couple from an established source, phdcomics - enjoy!
PS: I'm actually waiting for JeffE to comment on this: His credentials are such that I'm not even qualified to state them, and he himself claims that he had the lowest undergraduate GPA amongst any professor he's ever met!
EDIT: I'm talking about grades in subjects that you have to take as part of requirements of grad school - you would be expected to master the subjects that are directly required in your Ph.D research thesis, so your adviser would expect great grades in them naturally!
tl;dr - They don't matter (as long as you clear all your subjects!)
To be a bit blithe about it, grades matter until they don't. When you are a PhD student, the grades will matter until after you've completed your "qualification" process, in whatever form that takes. If you do well on the exams, then your grades don't matter much; if you're "on the bubble," you might be helped by solid performance in your graduate coursework.
Where grades continue to matter are:
- if you decide to apply for a graduate fellowship, in which case review committees will usually want to see evidence of strong academic performance at both the undergraduate and graduate levels.
- when you apply for jobs, as employers will similarly want to see evidence that you took coursework somewhat seriously. (Some employers may—wrongly, in my view—have GPA cutoffs for graduate students!)
Just to note one place where they do matter: they may help you find a willing advisor. I am not inclined to take on students who don't do outstanding work in the first-year graduate course that I teach. On the other hand, students who are at the top of the class may find faculty trying to attract them.
I currently direct a graduate program and as others have said, grades are most important while you're in a graduate program when decisions about funding are being made. C's may get you dismissed, and B's could lead to no funding in programs where funding is limited to a handful of students. If you are in a terminal MA program and applying to PhD programs, grades matter. After graduation and when you're on the job market, grades won't matter as much as your dissertation, publications and recommendations. The way that grades would matter then would be what they say about your professors' estimation of your work and whether they are likely to write you strong rec. letters. In short - if you get anything below an A, talk to the professor about what you could have done to do better and strive to get there.
I will actually take a strong opinion that my subpar grades from undergrad and graduate school are a large factor in why my PhD is going to take a long time. As David Ketcheson mentioned, quality grades and GRE scores are important denominators for Graduate fellowships. So far the NSF, NIH, and DOE have all denied me fellowships because I have a lowish GPA according to my reviews.
That being said, once you are an established researcher with a paper track record, the grades really shouldn't matter asides being a bragging point.
It depends, but what I have seen so far (w.r.t. Computer Science PhD perspective) is a bit different than other opinions in this post.
I am doing PhD (Computer Science) in the US (at a mediocre school) and most of my cohorts (not most, actually all of them) have a very decent GPA (like 3.9+ out of 4.0), even couple of them have perfect 4.0/4.0 and a very few of them are publishing pretty much good level of research works.
Moreover, most of my faculty members in my school have similar credentials (good grade, good research background, and most of them are from top notch schools like UMD, UPenn, UMich, VTech) -- although my school ranks 60+ in the nation (according to US News, in CS specialization).
Even PhD students in my department who did summer research intern in big companies like google, IBM, Microsoft are also equipped with a good GPA. Even some of them have no publication at all !!, even after being a PhD student for more than 2 years. You will be surprised to know that one of them is currently working as a research intern position at google with no research paper at all, he has been a PhD student for last 2.5 years !! Moreover, if you are from a good school, things will be lot easier.
So, what I see, situation is totally opposite in US. Here Pedigree > GPA > publications.
I used to be the "odd one out", I had a decent publication record (published 4 conference papers during my masters (from a top school outside US) and 2 more in my 1st year of PhD here), but I settled with a low grade (3.2+), compared to others it is actually a "bad". Moreover, the venues that I have published my research works are considered to be quite decent in my field (two of the venues are in the top 20 lists in Microsoft academic search ranking, even I got one best student paper award there. In the US, keeping up both research works and regular course load is quite challenging, at least for me.
Such situation compelled me to focus more on the grades and to start practicing problems from sites like "careercup" to land a job in industry -- that's what others are doing here, and this is a totally disappointing and frustrating experience for me.
I had to cut my time on research as much as possible, now just doing a minimal possible work to maintain my supervisor's "happiness threshold level". Here, landing an academic job is harder than in industries -- unless you are a "rock star" researcher -- and this is the reality. Being a mediocre researcher (like me) from a mediocre school does not count much, neither in industry nor academia.
Last but not the least, I am an international student deceived by the so called delusion of "the land of opportunity".
You may need good grades during the first two years of your PhD (the time you would be doing your MS, if you applied to a separate MS program) for the reasons listed above: fellowships, in case you decide to drop out and get a industry job, etc.
After two years in a PhD program, worrying about your GPA is a complete waste of time. I've yet to see any research university even ask for a transcript of your graduate classwork. Some research positions in industry might require a transcript and take into account your GPA, but I don't think this (misguided) practice is common.
What's true for certain is that if you're spending late nights and long hours polishing that final project in a class totally unrelated to your research area, so that you will get an A+ instead of an A -- please stop. You are hurting your career.
I am a Ph.D. student in my second year of research. I hold two MBA's and a bachelors degree. Each of my GPA's raised a bit in the course of my academic career (3.52 - 4.0). I am currently on the President's lists for Capella University. I do not think it matters to most. It certainly feels good to place a 4.0 in my application to teach at universities but I doubt it does much for my selection. The larger impact is what differences can you make. I have saught after a reputation to help develop peoples motivation towards academics, so a strong GPA helps prove that I know how to provide good advice on study habits to help others navigate through Bloom's Taxonomy. I would say, simply do your best! If you are in a Ph.D. program, you pretty much have a focused goal, and all the other crap is noise. Do your best, clear your classes, and focus on what is most important to you, not on what you think others will expect of you. Remember, Doctoral Students only account for less than 2% of America's population so take that into consideration when concerning yourself of what people will think of you. Be brave, take chances and have fun with your studies. You may lose points for being late, may lose points for a paper being too long, or perhaps you explored something not related to the assignment but got a lot out of it, don't stress about a perfect grade. A perfect GPA means nothing if you didn't maximize the opportunity to learn.