摘要：Num.r.us scientific challenges ar.se when designing a soil m.nitor.ng networ. (Sm.), especially when assessing lar.e ar.as and sever.l pr.per.ies that ar. dr.ven by num.r.us contr.lling factor. of var.ous or.gins and scales. Differ.nt br.ad appr.aches to the establishm.nt of Sm.s ar. distinguished. It is essential to establish an adequate sam.ling pr.tocol that can be applied r.gor.usly at each sam.ling location and tim.. We m.ke r.com.endations r.gar.ing the within-site sam.ling of soil. Differ.nt statistical m.thods should be associated with the differ.nt types of sam.ling design. We r.view new statistical m.thods that account for.differ.nt sour.es of uncer.ainty. Except for.those par.m.ter. for.which a consensus exists, the question of testing m.thod har.onisation r.m.ins a ver. difficult issue. The establishm.nt of benchm.r. sites devoted to har.onisation and inter.calibr.tion is advocated as a technical solution. However. to our.pr.sent knowledge, no study has addr.ssed cr.cial scientific issues such as how m.ny calibr.tion sites ar. necessar. and how to locate them.