To assess soil quality, a Minimum Data Set (MDS) of soil properties has to be proposed commonly through calculating the total load of each candidate soil parameters on all of the qualified principal components by the use of Principal Component Analysis (PCA) and Norm-value computation. Considering intensive land-use change, the method introduced in this study for MDS establishment integrated the quantified contribution of landuse forms and duration to each soil parameter using multivariate analysis and mean multiple comparison. In this way, a MDS representing maximally all candidates with minimal loss of the soil quality information contained by those non-MDS soil parameters is established. The MDS proposed can not only well integrate the quantified influence of land-use change and duration on soil parameters, but also is quite flexible and extendable with the potential to be extrapolated to assess soil quality in other regions. Based on two sets of soil database obtained separately in 1985 and 2004, two MDSs established, were compared with each other. It was found only a quite small change in MDS component occurred during a 20-year period. For a better assessment of soil quality, it seemed necessary to examine at what temporal scale and how much the MDS will change for a site-specific area with intensive landuse change.