OK, I agree these are legitimate questions. However, these comments definitely are the last for this thread.
p24 antigen usually appears in the blood (in measurable quantities) 1-2 weeks after catching HIV, rarely up to 3-4 weeks. As antibody is produced, it attacks p24 and removes it from the blood; as antibody levels rise, p24 levels decline. After a while -- usually 2-4 weeks, maybe up to 6-8 weeks after exposure -- detectable p24 disappears from the blood.
By 4 weeks the blood of everybody with a new HIV infection has measurable p24, antibody, or both; here are virtually no exceptions: either p24 or antibody will be detected. Some people may have both antibody and p24 antigen for several weeks, but eventually p24 disappears altogether. But for diagnostic purposes this doesn't matter: a negative combination test at 4 weeks or greater shows the person isn't infected.
Yesterday I posted a follow-up comment an another thread that goes into more detail about why some authorities continue to advise up to 3 months for definitive testing, despite newer tests that shorten the window period as well as evidence that even antibody-only tests are nearly 100% reliable by 6-8 weeks. I hope everybody reads it:
http://www.medhelp.org/posts/HIV-Prevention/-A-Question-on-Testing/show/1347755