A modified procedure for the determination of amylose in cereal starches and flours based on complex formation between the lectin concanavalin A (Con A) and amylopectin has been developed and characterised. The assay format is suitable for multi-sample analysis, allowing the analysis of up to 20 samples per day. In the procedure, the amylopectin in a solubilised, lipid-free starch sample is precipitated by reaction with Con A and removed by centrifugation. The amylose remaining in the supernatant is then determined after amylolytic hydrolysis to glucose and expressed as a proportion (%) of the glucose derived from amylolytic hydrolysis of the total starch in a separate aliquot of the solubilised sample (i.e. prior to Con A treatment). The Con A procedure correlates well (r>0·993) with existing Con A-based and iodine-based procedures and yields a linear standard curve for starch samples containing from ∼0 to ∼100% amylose. Advantages of this modified Con A procedure for amylose determination include its applicability to flour samples without the need for prior starch purification; it allows the simultaneous estimation of total starch and does not require a calibration curve. Repeated analyses of a set of samples yielded repeatability (within laboratory) relative standard deviations of