In　monitoring　and　controlling　wastewater　treatment　processes,　on-line　information　of　some　essential　wastewater　parameters　is　inaccessible.　In　order　to　acquire　the　real　time　data　and　contribute　to　real　time　closed　loop　control,　a　soft-measuring　approach　applied　in　wastewater　quality　measurement　is　put　forward　based　on　PCA　time-delay　neural　network.　It　is　composed　of　three　elements:　principal　components　analysis,　time-delay　neural　network　and　model　updating,　where　the　offline　model　is　trained　through　the　algorithm　GABP.　This　model,　which　is　of　good　real-time　property,　good　stability,　high　precision　and　easy　updating,　can　be　applied　to　on-line　predict　wastewater　BOD.