Abstract: The main objective of this paper is to develop a framework for forecasting of India’s annual exports at regular intervals, which would be carried out for principal trading partners and their principal commodities. Individual country/commodity analysis takes into account the country/commodity wise characteristics such as non-tariff barriers, language differences, locational/distance differences, preferential and other trading arrangements etc. Apart from the above mentioned country/commodity specific characteristics it may be due to the demand conditions, differences in the degree of the sensitiveness of prices, which cannot be captured at an aggregate level. The general framework of the econometric model for forecasting of India’s export by selected destinations and its selected commodities at the 6-digit HS classification level is developed in the first half of the paper. The econometric analysis of the framework rests on panel data. The primary factors taken to be influencing India’s exports at a disaggregated level are total imports of the destination country and relative prices reflecting the level of competitiveness of India’s exports in the destination country. The entire model constitutes of around 280 variables for the purpose of forecasting. For illustration purposes, the estimation of econometric sub-model for India’s export to USA is stated. There are 17 cross-sectional commodity codes with the time–series ranging from 1993-2001. The independent variables forecasting is performed using time-series models. The revised forecasts estimate for India’s export to USA stands at 8.85 per cent for 2003-04 with the base year of Mar02-Feb03.