Soil salinity is widely reported to be a major agricultural problem, p
articularly in irrigated agriculture, and research on salinity in plan
ts has produced a vast literature. However, there are only a handful o
f instances where cultivars have been developed which are resistant to
saline soils. Reasons for the lack of success in developing salt-resi
stant genotypes, and for the low impact that plant physiological resea
rch has made, are explored. We conclude that soil salinity has not yet
become a sufficient agricultural problem, other than on a local scale
, to make salt resistance a high priority objective for plant breeders
. The limited success of simple selection, where this has been practis
ed in breeding programs, can be accounted for by the fact that researc
h has consistently shown salt resistance is a complex character contro
lled by a number of genes or groups of genes and involves a number of
component traits which are likely to be quantitative in nature. We als
o conclude that the results of physiological research have been poorly
marketed by physiologists and, understandably, have failed to impress
plant breeders. We anticipate that the importance of salinity as a br
eeding objective will increase in the future. Our assessment of report
s of the degradation of irrigation systems, together with projections
of the future demands of irrigated agriculture, is that enhancing the
salt resistance of at least some crops will be necessary. Salinity res
istance will both help provide stability of yield in subsistence agric
ulture and, through moderating inputs, help limit salinisation in irri
gation systems with inadequate drainage. It is emphasised that plant i
mprovement and drainage engineering should be seen as partners and not
alternatives. We conclude with a personal view of one way forward for
developing salt-resistant genotypes, through the pyramiding of physio
logical characters.