Myth #1: You have to do a postdoc.
Negative, ghost rider. As a grad student nearing the end of your PhD work, you adviser will almost invariably talk to you about looking for a postdoc position--and from my colleagues' and my own experience, s/he will rarely mention other options. I suspect the reason behind this is that (a) it's what they know because it's what they did and/or (b) they assume all grad students are planning to head down the research track, if not the tenure track. If you have a pretty good idea of the path you want to forge, and you know it isn't the tenure track, then it's worth questioning this dogma. Talk to people in the positions you want to pursue; find out how important doing a postdoc is-or isn't.
Myth #2: There is only one type of postdoc-the academic research postdoc.
Most likely, when you hear the word "postdoc", you think of a PhD working at the bench in a lab with other postdocs and grad students at a research university. However, there are several different flavors of postdoc. You can do a research postdoc in a government lab or with a biotech or pharma company. If you're interested in teaching, there are postdoc fellowships and programs that mix teaching and research or focus on science education instead of research. You can even do a research postdoc at a predominantly undergrad institute; this grants the advantage of working with undergrads on a day-to-day basis. Plus some PUI departments, instead of hiring an adjunct for a semester, first offer postdocs the opportunity to fill open courses. I will confess that I know little about these other types of postdoc positions, as I am the typical research postdoc, but they do exist.
Myth #3: The only reason to do a research postdoc is to take a crack at the tenure track.
Another confession: I once thought that the only sensible reason for doing a postdoc was if you wanted to stay in academia and start your own lab. Sure, some people changed their mind along the way, but what was the point of doing a postdoc if you knew, upfront, that you didn't want your own lab? Turns out this is an absolutely ridiculous view. There are actually several career paths for which a research postdoc is preferred, if not required. Over lunch with trainees, the executive editor of a glamor mag (the science type, not the fashion type) commented that when hiring new editors, they liked seeing postdoc experience, especially in a field different from the applicant's PhD work. The reasoning, I gather, is that an editor should have some idea of how science is done and should be broadly trained, as a vast array of topics will be crossing his/her desk. So you go do a research postdoc, and after a year or two, you decide you're done with bench work; there are many places, both in and outside of academia, where that postdoc might help you get the job you really want.
These are major preconceptions that affect how grad students, postdocs, and advisers view and approach the postdoc. If they continue to be perpetuated, then it is going to be tough to have a productive conversation on the subject. What other urban legends about the postdoc need to be put to rest?