The concentrations of minerals and trace elements in coffee reflect the soil type and the growing conditions.Consequently, there is a pressing need for analytical methodologies to accurately authenticate the origin of coffee beans.This study was aimed to determine the geographical origin of green coffee beans from selected areas of Amhara region by using multi elemental analysis. Fourty green coffee bean samples, collected from four major coffee caltivating areas of the Amhara regionnamely West Gojjam, East Gojjam, Awi and Bahir Dar special zones were analyzed for the determination of K, Mg, Ca, Mn, Fe, Cu, Fe, Co, Ni, Zn, Si, Cr, Cd and Pb using inductively coupled plasma–optical emission spectroscopy. Different digestion procedures were tested by varying reagent volumes, time of digestion, temperature of digestion and types of reagents to develop a procedure that consume less reagent volume, short digestion time and low temperature for digestion. 0.5g of coffee sample was digested by the optimized condition. Differentconcentrations of metals were prepared by employing a six point calibration curve and good correlation coefficient was obtained, R2 >999 except for Si, Kand Mg that is 0.997, 0.998 and 0.998, respectively. Quantitative contribution of macroelements decreased in the order: K > Mg > Ca > Na, in all coffee assortments. Contribution of micro-elements those are essential bio-elements for our body decreased in the order: Fe > Mn > Cu > Zn > Ni > Co.here was a significant differences (α<0.05) between zones and districts.Principal component analysis was used to group the samples based on their elemental composition,72.2 % between groups and 11.5 % within groups. Application of linear discriminant analysis with the elemental concentrations resulted in an overall 91.7% correct classification of the coffee beans into their respective regions. The metal contents of the study may be taken as one possible tool for authentication of the regions coffee. Keywords: ICP-OES, Green Coffee Beans, Metals, Geographical Origin, Linear Discriminate