The main purpose oflhis sludy is 10 examine the relationship of FDIflow 10 Ethiopia wilhrneconomic growlh, expoI'I bOlh al aggregate and secloral level (manufacluring andrnagriculiure).rnA simple descriplive analysis, regression, causality, and correlalion analysisesrnlechniques were employed. Dala over 30 years (1981-2010) was usedfor regression andrncausality analysis. For secloral level analysises data over 1993-2010 was used. Thernregression equal ion is modeled wilh RGDP as dependent variable and FDI stock/flolVrnentered as independent variable of inleresl wilh olher control variables (domeslicrninvestment, labor force, lotal government expenditure, and inflation rate).rnThe largest share of FDI flow to Ethiopia is found to be in manufacluring seclor lvhichrnaccounts about j2 % and Ihe rest 22% and 26% goes to Ihe agriculture and servicernsector respeclively. Regionally, il is characrerized by uneven dislribution, Ihough Ihere isrna differentiated incentive slrategy to encourage investment in leasl developed regions.rnThe long run regression result shows that FDI stock is found to be positively relaled tornRGDP where as FDI flow is negatively related. In causality analysis, the Granger norncausality test failed to reject the hypothesis thai FDI stock does not granger causesrnRGDP. Conversely, the hypothesis that RGDP does not granger causes FDf slock isrnrejecled at j% significance level implying Ihe direction of causality ji'om economicrngrowth to FDI slock. In case of aggregate export and FDI stock, the null hypothesis thatrnboth variable does not granger cause each other is rejected al j% significance levelrnimplying bidirectional causalily between the variables. In examining Ihe relationshiprnbetween sectoral FD1 flow and sectoral export share as percenlage of merchandisernexport, for agriculture it is found to be moderate where as for manufacturing it is quiternlow