People often says that plasmids become unstable if you use them to clone long DNA fragments (over 20 kb), and that for this purpose is better to use cosmids (can carry DNA fragments up to 50 kb). However, as far as I know the only difference between both vectors is the presence of cos sites in the latter, and that those cos sites are important for DNA packaging into viral particles (not involved in vector replication). Considering that, how can we explain that cosmids remain stable with larger DNA segments?. To bring up an example, pHC79 is a cosmid derived from pBR322 plasmid (they are exactly the same except for the presence of lambda cos sites). pHC79 has been successfully used to clone DNA segments up to 50 kb. The question is: Can we expect that pBR322 works as well as pHC79?