The estimates in the government-funded surveys of the handloom and powerloom sectors of the number of units, looms and employment are not reliable since they are especially designed to identify units eligible for sector-specific schemes. The unit-wise Annual Survey of Industries and the National Sample Survey Organisation 62nd round data for 2005-06 cover the entire manufacturing data, but the results derived for variables such as value added, output, fabric production, etc, are also not considered reliable as the small units have a tendency to under-report their production in order to retain the benefits of small-scale sector tax incentives. This paper presents an analysis of the National Council of Applied Economic Research sample survey, which was designed to estimate the weighted per unit ratios for various size classes of units belonging to the handloom, powerloom/mill and knitting sectors.