The effects of windbreaks on pastures are reviewed, with an emphasis o
n temperate grazing systems. Mechanisms of plant response to shelter a
re dealt with in brief. Few papers on measured responses of pasture sp
ecies to shelter were located in a search of the global literature for
the period 1972-97. Except in cold climates, where the benefits of sn
ow-trapping on water availability can be demonstrated, there were few
reports of increased production of pasture in response to shelter. A s
ignificant result was obtained in a summer rainfall environment in Aus
tralia, where a 43% increase in wool production was obtained over thre
e years in small plots sheltered with iron sheeting on the fences. The
gain was attributed to increased pasture growth. In New Zealand, one
study over three years with a narrow, permeable shelter-belt in a wind
y, dry summer environment showed a 60% increase in pasture growth in t
he sheltered zone. However, another study on a high rainfall site with
a dense, wide shelter-belt found no substantial shelter effect on pas
ture. In dry, hot and windy climates there appears to be scope for pro
tecting spray-irrigated pasture with windbreaks. The feasibility of ev
aluating shelter effects on pastures or crops from old windbreaks is q
uestioned. Variability of soil over the site can not be satisfactorily
accounted for and there are problems in defining the true 'unsheltere
d' yield. Shelter effects on pastures could best be determined by comp
aring production in small completely sheltered plots and open plots. E
ffects in and near the competitive zone should be measured for living
windbreaks. Modelling could then be used to evaluate windbreak systems
. We are not yet in a position to provide unequivocal advice to farmer
s on windbreak outcomes for particular purposes or regions.