Reactor thermal hydraulic research began together with the first light
water reactor early in the sixties. Extensive experimental research p
rograms have been carried out to increase understanding of the behavio
r of light water reactors under accident conditions and to validate th
e computer code developed. An example of the developmental stages for
a system code is given. Problems associated with the use of system cod
es are identified and possible solutions are given. The application of
system codes in the design and licensing of innovative reactors and t
he development and qualification of uncertainty methodologies, constit
ute issues of current interest in this area of research that are consi
dered in the present paper. Perspectives and requirements for future r
esearch are outlined.