Using multi-temporal NOAA-AVHRRdata, combine with historical data, this paper deals with how to use NDVIto monitor and forecast pest which mainly caused by pine caterpillar moth (Dendrolimus punctatus walker). First, geometric correction and registration for the images were made; Second, through extracting image data and rejecting cloud contamination pixels in monitor regions, all monitor regions were separated into different districts according to the features of geography and climate, then every district was separated into pest injury and non-injury areas; At last, NDVIvalues of every pixels were computed, the statistic values about two areas were computed and time-series curves of NDVIstatistics about two areas were compared. It was discovered that time-series curves of NDVIaverage value can be used to monitor forest pest happening, time-series curves of NDVIvariation coefficient can be used to forecast. In the end, the paper provids the prospect of using NOAA-AVHRRimages to monitor and forecast forest pest injury.