Microwave remote sensing provides a unique capability for direct obser
vation of soil moisture. Remote measurements from space afford the pos
sibility of obtaining frequent, global sampling of soil moisture over
a large fraction of the Earth's land surface. Microwave measurements h
ave the benefit of being largely unaffected by cloud cover and variabl
e surface solar illumination, but accurate soil moisture estimates are
limited to regions that have either bare soil or low to moderate amou
nts of vegetation cover. A particular advantage of passive microwave s
ensors is that in the absence of significant vegetation cover soil moi
sture is the dominant effect on the received signal. The spatial resol
utions of passive microwave soil moisture sensors currently considered
for space operation are in the range 10-20 km. The most useful freque
ncy range for soil moisture sensing is 1-5 GHz. System design consider
ations include optimum choice of frequencies, polarizations, and scann
ing configurations, based on trade-offs between requirements for high
vegetation penetration capability, freedom from electromagnetic interf
erence, manageable antenna size and complexity, and the requirement th
at a sufficient number of information channels be available to correct
for perturbing geophysical effects. This paper outlines the basic pri
nciples of the passive microwave technique for soil moisture sensing,
and reviews briefly the status of current retrieval methods. Particula
rly promising are methods for optimally assimilating passive microwave
data into hydrologic models. Further studies are needed to investigat
e the effects on microwave observations of within-footprint spatial he
terogeneity of vegetation cover and subsurface soil characteristics, a
nd to assess the limitations imposed by heterogeneity on the retrievab
ility of large-scale soil moisture information from remote observation
s.