Just because Poser shows more memory consumed after being used and then reset to "new" doesn't make it a leak. If they are being smart and delay-loading libraries (and I mean DLLs), then those won't be in-memory until they get used the first time. There are other legitimate reasons Poser may consume that memory without it being a leak. Now, if subsequent File->New operations consume steadily more memory, that is good evidence of a leak. Is that what you see happen? By the way, Task Manager is a terrible thing to try to accurately track memory leaks with. To demonstrate this, take your "leaking" scenario and then minimize Poser. Task Manager will claim that Poser just gave back a bunch of memory, which it most certainly didn't. Perfmon is a better tool for this kind of tracking, though it is rather harder to use.