This paper reviews methods that have and can be used to forecast the effect
of changes in accessibility to the rail network on the demand for inter-ur
ban rail travel and of available evidence on rail accessibility elasticitie
s. It reveals that relatively little research has been conducted in this ar
ea and that the forecasting procedures that could be used imply large varia
tion im accessibility elasticities, which has not been empirically justifie
d. Fresh empirical evidence on two related matters is reported. First, the
neglected area of choice set composition is examined and the extent to whic
h rail is considered to be a realistic alternative for inter-urban journeys
and the contribution that accessibility to the rail network makes to this
are analysed. Second, rail trip rate models are presented that not only con
tain estimates of accessibility elasticities and of the effects of a range
of socio-economic variables on the demand for rail travel, but also that al
low tests of the accessibility elasticity variation implied by many forecas
ting procedures to be conducted. It is found that this elasticity variation
is not empirically supported. It is concluded that there is only limited s
cope for increasing rail demand through improvements in rail network access
ibility.