Recently some networks for monitoring the soil quality have enriched t
he world of environmental quality assessment. Formerly such networks w
ere considered not worthwhile because the spatial variability of the s
oil compartment was regarded too large to detect trends in soil qualit
y. However, research suggested that with a statistically sound network
design it should be possible to detect relevant trends. As a result,
national and provincial. authorities in the Netherlands began to set u
p soil monitoring networks in the early 1990s. These networks cover th
e rural areas and mainly focus on three environmental aspects of the s
oil, namely: spread of contaminants (heavy metals, pesticides, PCB, et
c.), and eutrophication and acidification. The goals are the assessmen
t of the current soil quality of rural areas and the development of th
is quality. An evaluation of the first 5 years of monitoring yields mu
ch useful information. Naturally, after the evaluation of the first re
sults, optimization of the organization and design of these networks m
ust be considered. Topics discussed in this paper include sampling tec
hnique, sampling design and the maintenance oi sample archives. Topics
of a different category are the concept of homogeneous subregions, th
e idea of fixed monitoring locations versus a flexible network, season
al variation and its influence on monitoring, and finally the choice o
f measured parameters. (C) 1998 Elsevier Science B.V. All rights reser
ved.