The purpose of this study was to develop operationally important soil quality indicators to evaluate long-term sustainability, at the farm scale, for no-tillage systems in Argiudolls of rolling pampa (Argentina). The soil was classified as series Arroyo Dulce (Typic Argiudoll), a fertile dark, deep and well-drained soil of the hills. Three situations were considered: pristine soil with grass vegetation, grassland soil (also considered as a reference situation); and 15 years no-tillage soils from four production plots. Physical, physico-chemical, chemical and biochemical indicators were considered. Data were analyzed by principal components analysis (PCA) with canonical discriminant analysis (CDA). The first three components explained 90% of the overall variation. For pristine undisturbed soil, the main variables selected by PCA were particulate C, pH, respiration and total organic C, and in the case of grassland they were C stock (mass of C in the 0-10 cm soil horizon), water-soluble C, and % silt. The no-tillage area was separated in different plots according to the degree of erosion with different depths of the A horizon. Clay content and bulk density were the main variables in the less degraded no tillage plots. Cluster analysis was applied to construct an average linkage distance dendrogram.