This paper examines the salient characteristics of regional structure of Indian agriculture system after the introduction of liberalisation policies in the early 1990s. At the backdrop of physiographic regions, district has been used as an Operational Taxonomic Unit (OUT) to analyse the inter- and intra-regional variability of Indian agriculture system. Intra-regional variation has been analysed by amalgamating geographic space into similar OTUs within each physiographic region. A “cluster combination technique” is applied for the identification of homogenous regions using a number of variables related to productivity levels, inputs, demographic attributes and infrastructure characteristics to represent transformation in the post-liberalisation period (2000-2001). After the process of amalgamation of OTUs into regions for each physiographic region, it is concluded that there is an effect of agglomeration economies on intra- regional variability particularly in regions located in the Great Plains of India. The size and growth of local towns and the density of road network also have a direct bearing on changes in regional structure of Indian agricultural system.