At a company I used to work for, I got in trouble with my supervisor and the CEO upon completion of a huge project on which I was the lead developer. When we first defined the scope of the project, my immediate supervisor asked me about how many programmer-hours it would take to complete the project. After a day of going over the spec with the programmer who would be overseeing a major component of the project (I'd be building the other major component myself), I said it would take about 1400 programmer-hours.
My supervisor assumed that the 1400 hours figure was a complete BS answer, and that no project could actually take that long to build, so he lowballed it in his estimate to the higher ups as 1000 hours, figuring that his stellar management skills (that is, calling us away from working to sit in meetings to talk about why the project was so massive) would help trim the time down. Of course, my figure included assumptions that the client would surprise us with unexpected data flow and that we'd waste a whole lot of hours in meetings.
The week after the project launched, when the timesheets were compiled, it turned out that the project had taken 1375 programmer-hours to complete. My immediate supervisor was furious, because he honestly felt that I'd been kidding about the 1400 hour figure, even though I'd provided solid documentation of the breakdown of the work. (It didn't help that he didn't actually understand the project, or why it was complicated.)
That next week, the CEO gave me an hour long reaming about the project. Apparently, the company lost a huge amount of money on it. It seems that my immediate boss had reported to his superiors that his estimate of 1000 hours was based on applying a 15% overrun to my estimate of how long it would take, and that there was no way we would take longer than 1000 hours. The company had then based the price quoted to the client on the 1000 hour figure, with a reasonable profit margin on top. Of course, the 1375 hours spent on the project was well above the break-even point on what should have been a hugely profitable project for the company, all because my supervisor didn't like my estimate...