I propose that the very earliest cells could not have contained intrac
ellular parasites in the sense of independent, self-benefiting agents
that were detrimental to their hosts. Before there were effective ways
to extract energy by methanogenesis and photosynthesis, the world eco
system, I have argued, was very sparse and essentially monophyletic (K
och, 1994). During much or all of this time, nontransmissible agents,
comparable to plasmids, would have been selected against as long as th
ey could not synchronize their growth with that of their single-celled
host. Even when such an adaptation was initiated, it must have been o
ne that was likely crude and imperfect, and frequently failed. The tra
nsmission of genetic material via subcellular vectors would have depen
ded on the later development of special mechanisms for the introductio
n of nucleic acids into living cells. Only after living forms became a
bundant could an efficient and effective transmissible intracellular p
athogen prosper, because by then a transmissible agent could be destru
ctive to its host and move successfully to a new host. The implication
is that much of the earliest evolution of either the host's cellular
functions or the parasite's techniques and strategies proceeded withou
t gene transfer. Support for these ideas is presented, starting from h
ypotheses concerning the nature of the first life forms and leading to
a scenario for the development of transmissible intracellular pathoge
ns.