For my professorial readers (or otherwise readers with PI status):
How long does it take you to write, from scratch, a single-investigator grant proposal that will undergo peer review?
For physical scientists and many others, I am talking about a standard NSF 15-pager or similar. For biomedical folks, that would be an NIH R01 (however long it is).
I am assuming we are talking about a new proposal (not a resubmission), and I am also assuming that you don’t just drop everything and disappear into a cave to write, but rather continue to tend to other normal academic duties, like teaching or service.
xykademiqz 
I voted 4-6 weeks, but as a caveat, I believe that all the R01s I’ve submitted for many years have been at least partial re-workings of shorter grants submitted to other mechanisms – foundations, internal opportunities, etc. It has been 8 years since I have written an R01 100% from scratch, and I’m pretty sure that one took more than 8 weeks, but I was a newbie then …
Ideally (8 weeks or longer). in actual practice (4-6 weeks).
I can write one in a week, but I do need to disappear into a cave. 8 weeks or longer is ideal, but I am an incurable procrastinator and usually only have less than a few weeks left, so the cave becomes my only option. I’ll get started on my NSF proposal later today (unfortunately, I am not kidding here).
I put in 3 weeks, that is the actual writing time from a rough outline to submission. However, the ideas for these things are formed over months and I keep track of things in a notebook. So maybe that is why writing is ‘shortish’. I do tend to drop non essential things during this period (like revising a paper in detail for journal submission).
I am working on two NSF proposals to submit this cycle, with about 8 weeks total for both. I have been working on the first one for 3 weeks now and have given myself till mid next week to finish it, then on to the next one (they are very different, going to different directorates). I am also teaching a new undergrad class and have the usual service. With 4 weeks per proposal, I have I have been staying late (till 8-8:30) most evenings this week and this will probably continue till the deadlines in early November (DH has been a good sport about it); so 4 weeks but with extra 10-15 extra hours of work per week. I think some degree of being in a “cave” is necessary, at least for me, because I seem to be most inspired to write compelling prose (the intrigue! the excitement!) in late afternoons/early evenings. The mid-afternoon coffee is very invigorating!
I just banged out a very solid new R01 from scratch in two weeks while submitting two manuscripts at the same time.
How often can you submit to the NIH? Does it depend on the institute?
The NSF folks are very nervous these days since it’s once a year chance to submit across the NSF.
NIH has three submission deadlines per year and I don’t think there are any limits on how many proposals you can submit per year or per deadline.
It usually takes me about 4 weeks of pretty focused effort to put together a good quality R01 sized proposal. The first few attempts took much longer. I’m still a relative newbie but I’m getting a little faster with experience.
Suggestion for another poll, just for kicks: How long does it take to write up a manuscript, when *most* of the data analysis is done? Or, if you are a PI and haven’t done this in a while, how long did it take you when you first started out and had to do it all yourself? (Oh, and we’re talking society-level journal, not C/N/S.)
@Anonymous: How long does it take to write up a manuscript, when *most* of the data analysis is done?
For me this is about 3-4 weeks with other stuff going on, but can do it in 2 weeks, if I get into a cave and do nothing else but write. However, I find it’s better if I write it up, do one or two rounds of editing, then leave it for a few days or a week (so that I get some distance) and then do another final editing round.
Anon, I feel that’s a much more difficult question to answer. With grants, at least in the US, there are a handful of agencies that do peer review, so proposals have the same format across many disciplines. With papers, there are letter papers and comprehensive papers, there is theory versus experiment, and a whole continuum of prestige of journals. And that’s just in a single field!
For instance, in my physical science field, a society-level, 6-8 page, double-column paper can take as little as 2 weeks from start to submission, probably no more than 4 weeks (I am thinking Phys. Rev. B). Less time if I don’t teach, e.g. if I am writing in the summer. But a very long paper will take considerably longer; we recently submitted an 18-pager, it had something like 25 figures, and it took almost 2 months from start to submission. On the other hand, a letter paper for the very prestigious Phys Rev Lett, even though it’s 4 pages, has to be written ‘just so,’ so it easily takes several months for us because the key is to present the data in a very convincing and compelling fashion, which means only after you start writing do you realize you need more or different data and then go back to do more work. My experience with Nature and Nature Offshoot has been similar (collaborations with two other groups); easily a year or two from first draft to final submission, with more work done while the writing developed. (It doesn’t help that all prestigious journals with short papers now seem to expect 30-page supplements.)
I know this is completely out of whack, but I just found out that a student has copied my NSF proposal for his GRFP and got awarded the fellowship. What should I do? On one side I think it is only a student and I should let it go, on the other hand this level of dishonesty is unacceptable.
Sameir, this is a tough one. I will post your question and let’s see what the blogosphere advises.