The effects of several factors, like screw rpm, die restriction (11r) and moisture, together with corn endosperm hardness and rice amylose content, on melt viscosity, were analyzed using a Brabender single screw extruder. Bagley's method was utilized to estimate melt viscosity at the die and to obtain theological models. Relationships among different responses such as mass output, die pressure, specific mechanical energy consumption, melt viscosity and degree of cooking were discussed. All responses analyzed were found to be dependent on both extrusion conditions and material structural characteristics. Very low values of the exponent of the viscosity power law model "n" were found, particularly at low moisture. This fact is explained on the basis that to get different value pairs of mass output and pressure, rpm has to be changed and, as rpm increases, structural damage (degree of cooking) also increases, which in turn produces a reduction in pressure. It was concluded that "Bagley's method" should not be used to estimate melt viscosity in cases where pressure and mass output pairs of data obtained do not correspond to samples differing excessively in degree of cooking. (c) 2005 Elsevier Ltd. All rights reserved.