When submitting papers to journals, I understand from experience that acceptance can take quite some time (on the order of several months), even when a reviewer is immediately assigned to the paper. But can a referee usually tell fairly quickly if the article should be rejected? Are there any horror stories of waiting several months after a reviewer agrees to look at your paper, only to get a rejection letter?
The answer depends on the journal, the paper, and the reviewer(s). You might hear relatively quickly if there is a trusted reviewer available quickly and he/she isn't too busy with other work and the paper is either especially good, or especially bad. Often the bad is easier to judge than the good, of course. The longest reviews are the intermediate papers, especially if different reviewers have different opinions about it. "We want to publish this. Do we have room in an upcoming issue? We have a lot of competing papers. I love it. I don't think it's quite ripe yet. Etc. Etc. Etc."
Note also that the reviews of intermediate papers (and even some judged to be quite good) may come with some advice for improvement that will, perhaps, lead to acceptance after additional review. This will depend on the journal, of course.
However, don't wait to hear the judgement of the reviewers. Continue work on the project, working to improve both the result and its presentation. Work to extend the result.
Yes, it can happen. I've seen one paper which was rejected after four rounds of revisions. The reviewer was dubious about the method employed, but was not confident enough to advocate immediate rejection. After some back-and-forth with the authors the reviewer eventually decided that the method was too adventurous to give reliable results, and the paper was rejected.
Having said that, it's possible also that the reviewer writes a quick report with "this paper is so bad, it doesn't have X, Y and Z, and I'm shocked you're even considering publishing it". The exact specifics will depend on the paper in question.
Here's a source on the rough timeline in mathematics (not necessarily applicable to your field). To quote:
Our average time to reject a non-serious submission is 7 days, our average to reject a more serious submission is 47 days, and our average time to accept is 121 days. There is considerable variance in these figures, so they should be interpreted cautiously.
We once had a rejection after about 24 months and the journal would refuse to give us any meaningful status updates that would have helped up decide whether to withdraw the submission and go with another journal. On one occasion, after waiting for about 16 months to hear back about a submission, I wrote to the editors directly, politely telling them that I was considering withdrawing my paper with them. After a couple of weeks I have received the reports from the referees and the paper was accepted.
These situations are always a bit tricky to handle. If you decide to go with another journal, there is no guarantee that the process will be faster, and besides you have probably waited for some months already. Try to find out the typical wait period of a journal with your colleagues first to get an idea of what could be in store for you, keeping in mind that every submission is a story on its own.
I had a rejection once after 7 months of waiting accompanied with review reports that in my opinion did not motivate a rejection. I guess my paper was just felt out of scope for this journal.
I suspect there will be many horror stories. If you want to break the chain just continue (all work will eventually be accepted somewhere) and make sure you don’t become ‘such a reviewer’ later on in your career. As a reviewer, I fight against my own desire to take revenge when I notice thoughts crossing my mind like ‘my work was rejected for less’. Unfortunately, the review process lacks transparency and external quality control. There is no training or feedback for reviewers either.
Short Answer: Desk rejects are quick, but if it goes out to review, then rejections tend to take as long as any other first-round outcome. So if you haven't heard anything from a journal after 2 or more months, it probably means that your article has been sent out for review, but nothing more.
Desk reject: This is where the article does not get sent out for external review. Rather someone on the editorial team (e.g., editor, action editor, a team) review the manuscript and deem that it is not worth sending out for review. Desk rejects are usually quite quick (e.g., 1 to 4 weeks is common in my experience). Note that not all journals do desk rejects, and many journals vary in how much they filter at this stage.
First round rejections: This is where the paper is rejected after the first round of external review. In my experience, the time to reach a decision is often unrelated to the outcome, and therefore it takes as long as the review process takes. Your question implies that a reviewer looks at a paper and can quickly determine that it should be rejected. It's more likely that the manuscript is sitting in a reviewers inbox, and the reviewer is waiting to find the half-day or so to review the paper. In my area 2 to 4 months for an external review is common. But as you'll read, first round review times vary a lot between journals and fields, and from manuscript to manuscript.
That said, occasionally I've had review assignments taken away from me around the one month mark because one of the other reviewers gave a review sufficient to make it clear that the paper should be rejected. And sometimes editors will get an additional reviewer, if the first set of reviewers are inconclusive. So occasionally, the external review process may be quicker if it's a clear reject. But in general, the length of the first round review process is not diagnostic as to whether it will be accepted or rejected.
Second/third/etc round rejections: Your question does not seem to be asking about rejections at this stage. But the general point remains that the time to get a response is relatively unrelated to the outcome. Perhaps the main question here related to time wasting is where the journal gives you a revise and resubmit but then rejects your manuscript after you make the revisions. In general, good journals will give you a sense of what the likely outcome is if you conscientiously make the revisions. Some will use language like accept subject to revisions, minor revisions, major revisions, allowed to submit again, etc. This can give you some sense of the scale of revisions. Sometimes the editor will say that you can resubmit, but that it is a "high-risk resubmission". Thus, good journals will manage expectations about what the likely outcome will be given a thorough and conscientious implementation of the requested revisions.
Horror stories: You also asked the following question:
Are there any horror stories of waiting several months after a reviewer agrees to look at your paper, only to get a rejection letter?
First, time between submission and acceptance might look something like this:
As an author, you don't usually find out how long the components of the review process take. Your question implies that there is one person "a reviewer" deciding to review and then rejecting. But rather, these are usually different people, and there are multiple reviewers. At most journals, the external reviewers inform and make recommendations to an editor / action editor. They do not make the decision.
So in general, there are different norms about how long a review takes. But in my field of psychology (1 month is amazing, 2 months is nice, 3 months is par, 4 months is okay, 5-6 months is slow, beyond 6 months would be concerning). Other fields and journals have different norms. The point is I would not see waiting 4 to 6 months as a horror story. It's a bit slow. But It's just how long things sometimes take, and something that might factor into where I send my work in the future. And the length of time has almost nothing to do with the outcome. So in general, rejection is unpleasant and slow review processes are also unpleasant, but they're separate issues.
If the first-round review process is taking an amount of time well beyond the norms of your discipline, then you may wish to contact the journal to see what is the hold-up. Based on the response you get, you could weigh up the pros and cons of withdrawing your manuscript. That said, given the energy that may have already been invested in reviewing manuscript, I'd see withdrawing a manuscript as very much a last resort.
In my experience, rejections are only quick if the editor rejects it before sending to review. Now if a paper-to-be-rejected makes it to review, then a lot of funny things may happen. Low quality papers are a pain to review and it's common for reviewers to let them sit on the shelf for a while before they just make something up and send them back with "reject" label. That way it looks like the reviewer did some review work. Now if the reviewer is quick and sends the paper back to the editor same day saying "Pfft @#$%!", then the editor may just let it sit in his mailbox for a while because technically his journal provides high quality reviews and those can't be done in 15 minutes, can they? Besides, the editor certainly doesn't want to get any more papers from that guy. Better not get him excited with quick processing of papers here. He might even advertise it to his equally-"talented" friends. So 2-4 weeks seems like a reasonable time for rejection, unless the editor/reviewer forgets about it after all... But editors and reviewers never forget because they have nothing much else to do but review papers, do they? So here is a horror story for you :)
In my (pure math) experience, rejections after review are often significantly faster than other outcomes (minor corrections etc.). There are two likely reasons for this.