How seriously should a graduate student take teaching evaluations? Pretty much the question in the title.  If a grad student gets bad reviews as a TA, how much does that hurt them later? How much do good reviews help? What if the situation is more complex? (For instance, bad reviews when TAing, but good reviews when actually teaching/lecturing a summer course).
Edit: I asked this question with the situation of a student hoping for a career at research universities in mind, however, I am also interested in other cases.
Edit: In your answer, please mention what your background is: have you served on hiring committees?  Are you reporting just what you've heard?  Were you successful/unsuccessful in a job search and were told that your teaching evals did/did not make a difference?
 A: If the evaluations give articulate suggestions on how to improve, then, by all means, take the suggestions and try to improve. But don't try to change who you are to make yourself a different teacher. Some are impeccably organized. If being impeccably organized will make you feel dull, then don't do it. Do, however, indicate to your students that you have a plan or an outline in mind. Similarly, if you don't want to take road trips while you are teaching, then stay on the highway. Students can smell a fake. Be sincere, whether or not you mean it.
A: This is more of a comment than an answer, but I think it bears saying, based on some of the comments at the top.  Anyone who is now a math grad student was not a typical member of an undergrad math class.  While clearly the correlation between evaluations and good teaching is not perfect, it is not negligible either.  In my experience (which includes many years of recieving evaluations, at several different institutions, many years of reading teaching letters, and other related considerations of this question), if someone gets very low evaluations, there is typically a good reason for it.  While it may not be easy to get top evaluations, reasonable teaching almost always avoids terrible ones.  
And when one is recruiting post-docs, or more senior candidates, one wants to have some assurance that they won't be a teaching disaster.  If they are a brilliant teacher, all the better,
but one is more concerned about avoiding disasters, since these lead to trouble and difficulty for everyone.  
If you are applying to research positions, if your teaching is average or better, this should be fine.  If your evaluations are consistently and significantly below average, this is probably a sign that you need to put some effort into improving your teaching.  Students normally react well to clear explanation, sincere treatment, some humility on the part of the lecturer, and a little effort to be organized.  Negative evaluations often correlate with the violation of one or more of these principles.    Very good teaching does much more, but just doing a reasonable job should be enough to keep your evaluations reasonable.  
A: Positions not associated with teaching (such as industrial or government labs) will very rarely care about your teaching. When I applied for these, I didn't even bother to list my teaching on the cv.
Teaching positions (such as at a community college) will probably care about it a lot more, since they want some proof that you can teach well. But I can't say much since I don't have experience with these.
Research universities are somewhere in-between. In general, their main priority is the quality of your research. So for a standard tenure-track faculty positions, they will likely focus on selecting an interesting (research-wise) colleague rather than the best teacher.
Of course, research universities need to teach too, and they do feel the pressure to teach well. Also, "research university" is not a uniform designation; different universities will have different priorities which may include more or less emphasis on teaching.
Generally, teaching works like this at a research university. The department (math, in your case) needs to teach some courses. These are service courses to other departments (such as "calculus 1 for biology students") and internal courses (e.g. "graduate group theory"). These need to be taught adequately. If the service courses are not taught well, other departments will complain and your dean will not like it. If the internal courses are not taught well, then your colleagues will have underprepared students to deal with, and they will not like it. So people will want to know that you can teach adequately. Generally, at a research university, I would take "adequately" to mean that you will not leave the students grossly underprepared. Whether they love your teaching or not is less of an issue. So, as long as you have some teaching experience, I would say you are OK.
Now, you don't have to list ALL teaching evaluations on your cv. If the evaluations are great, mention them. If not, you can omit them and just list the course. For example:
TEACHING
Fall 2008: Calculus 1
Spring 2009: Algebra (received 4.5 / 5 evaluation)
Fall 2009: Linear algebra
...
Also, I don't think good evaluations will affect your candidacy negatively. It's true that some people might interpret interest in teaching as lack of interest in research, but I don't think good evaluations are enough for that. If you teach a lot, if you publish papers on teaching, go to teaching conferences, etc. -- in that case, yes, people might be suspicious of whether you are interested in research at all (especially if you don't have an equally active research program). But I don't think that just having good evaluations will do you any harm.
A: You should keep all your evaluations and keep track of them.
Are you improving over time ?
Are you improving in some areas over time ?
Are you better/worse in some topics than others ?
Take every comment with a grain of salt.
Some students will hate you and/or your teaching style no matter how you teach
either because they hate the topic, hate the class time, hate you, or hate your style
of teaching/grading.
[It is hard to give someone the highest possible evaluation when you get a 'D' in
the class.]
Look for patterns in the comments and accept that you cannot teach in such a way
to please all possible students.
Strive for clarity and helpful pacing.
Even if you end up at a Research University or Lab you will have to either teach
Lectures or Grad Seminars or give talks and the clearer and better paced they are
the better for everyone involved.
A: If you are a woman or a member of an "under-represented group" in mathematics, then there is some evidence to suggest that your teaching evaluations may be lower than they should be.  The literature is equivocal and you can do your own research by googling around on "implicit bias" and "student evaluation of teaching."  One principle seems to be that if you don't fit the implicit picture of a mathematician -- that is, you're not a geeky white male -- then implicit bias causes students to suspect your competence and rate you lower.
I've heard a variety of ways of counteracting this -- I've heard people urge women and members of other under-represented groups to dress more professionally in class than your typical white male professor does, so they are seen as more competent. For graduate students, having the professor who's teaching the class overall formally endorse your competence by introducing you at the beginning of the semester (this is so-and-so, a highly talented graduate student in mathematics with an undergraduate degree for XXX; she'll be running your discussion section) is probably a good idea. 
My own view is that if you are teaching and you sense from the atmosphere in the class that you're not getting the respect you are entitled to by virtue of your academic achievements, you should consult immediately with someone in power.  Don't let it get away from you.
A: The two universities for which I've read postdoc applications both require a teaching letter, so I've read a lot of these.  Two notes:


*

*Low evaluations the first few times teaching are quite normal.  If you do a better job later in your graduate career, early miscues tend to be forgiven.

*If you're concerned about poor evaluations (and even if you're not) you should ask whoever's writing your teaching letter to sit in on one of your classes.  I've read many teaching letters that say "such-and-such has mediocre teaching evaluations, which I think is due mostly to a lack of classroom pizzazz; I sat in on such-and-such's lecture, and it was very clear, and such-and-such did an excellent job understanding and responding to student questions."  This is much better than a letter that says "such-and-such has mediocre teaching evaluations but has not actually injured any undergraduates."

*I do think that "bad evaluations when grading, good evaluations when teaching" is a good combination: you're applying for jobs as a teacher, not as a grader.

*Of course, getting good evaluations and being a good teacher of undergraduates have something to do with each other but are not the same thing.  Some very simple ways, not a lot of work for you, to get better evaluations:  Learn as many students' names as you can.  Give the impression that you enjoy teaching the class, whether or not this is true.  (The easiest way to do this is to figure out which teaching assignments you'll enjoy most, and aim for those.)  Give the impression that you take the course seriously:  this should actually be true, but in case it's not, you should still try to give the impression that it is.  Will these things make you a better teacher?  I don't know, but I think they tend to create an atmosphere in which students are willing to work harder -- and this, after all, much more than the quality of your lectures, is what's going to determine how much math they learn.
A: You should worry about your teaching evaluations before you get them.  In other words, you should put effort into your teaching as 1) it does matter at almost every institution, and 2) you are teaching people math, and that is important.  It is also a good idea to discuss the evaluations with your students before you hand them out.  Remind them that this feedback will help you to improve your teaching so you would appreciate them telling you what worked and what didn't.  Often, only "angry" students will respond in the comment sections. By taking five minutes and reminding the students that the evaluations are a tool for communication, you should get much better feedback. 
As for worrying about them afterward, it is true that they are a secondary consideration at most research universities.  However, it is not helpful to have anything on your application that puts you lower on the list.  If your evaluations are bad, make sure that the person who writes your teaching letter speaks about the positive aspects of your teaching.  If you have to send in the evaluations, then have them address directly the mitigating factors (if there were any) that caused the bad evaluations.  
A: If you are viewed as a star research mathematician, I think it is safe to say that your student evaluations will do absolutely no damage to you at all.
If, however, you are viewed as anything less than a star and have to compete against many other candidates of comparable quality for a tenure track position at a school other than the top ranked schools, then the department will care a lot about whether you will be a good citizen of the department. This is only sensible, since if they eventually give you tenure, they're most likely stuck with you for life. Not provoking complaints by students to the university administration about your teaching is perhaps the most important aspect of being a good citizen.
A: There are two separate questions you might be asking here:
1) How much should one care about having bad ratings relative to average ratings
2) How much should one care about having average ratings relative to good ratings
I think in the first case you should care a lot more.  If your main goal is to get a job at a research institution then my guess is that there are more doors that close from being a bad teacher than there are doors that open by being a great teacher.  
Furthermore, not all aspects of teaching are reflected in ratings and you could certainly be a good teacher and get ratings that aren't great.  I'm mostly speaking from the experience of evaluating mentors at Mathcamp, where I have the most experience comparing student ratings to the independent evaluation of teaching quality by observers.  There are lots of other factors that influence ratings beyond teacher quality, and in the middle of the curve these other factors make a bigger difference than quality.  That said, both at Mathcamp and at Berkeley, it seemed the great teachers got great evaluations, and terrible teachers got terrible evaluations.  At Mathcamp when we had teachers who we thought were good who only got mediocre evaluations we were generally willing to overrule the students opinions, and that's in a situation where our primary concern is teaching.
A: I think that the "good evaluations will hurt your career as a researcher" meme is over-emphasized.  At almost every institution in the U.S., people are wary of hiring someone who isn't an adequate teacher: it always causes pain down the line for the hiree's colleagues,
as they have to deal with student complaints, possibly deans' complaints, and so on.
One should focus (at least) enough on teaching to do a reasonable job, and (hence) get reasonable evaluations.  Whether you want to focus more than that depends on your own personality, career goals, and so on.
A: I find it depressing and disconcerting that in the phrasing of your question, the subtextual modifier "with regard to future job prospects" was deemed so automatic as to make its explicit inclusion unnecessary.
A: I've heard everything from "without good teaching evaluations you will never get a job" to "good teaching evaluations will make people suspicious that you can't do research" -- so I don't suspect there's a consensus on this.
