In this paper attempts are made to identify determinants of kerosene and LPG demand inrnEthiopia. The study estimates kerosene and LPG demand using time series data staltingrn1964/65 to 2004/05 . To detemline whether there exists long run relation betweenrnkerosene and LPG demand and its detenninants, the johansen procedure of cointegrationrnanalysis used. Accordingly, there exists one unique co-integration relationrnbetween kerosene and LPG demand, and its result shows that the real price of kerosene,rnreal price of LPG, real per capita income and real foreign exchange eamings are found tornbe statistically significant while population growth and level of urbanization arernstatistically insignificant. Similarly in the case of LPG demand; real price of LPG (ownrnprice), real price of kerosene (relative price), population growth, level of urbanizationrnand real foreign exchange earnings are found to be statistically significant while only realrnper capita income is statistically insignificant. In the short-run, price of Kerosene, real perrncapita income and population growth are statistically significant in kerosene demand andrnreal price of LPG, real per capita income, population growth are statically significant inrnLPG demand. In terms of elasticities, kerosene price and LPG price are elastic andrninelastic, respectively while income is marginally elastic for both fuel demand model. Inrnthe â€¢long run, LPG is a substitute to Kerosene and electricity, other modern energy so urcesrnare a substitute to LPG ..rnfinally, in the short run, the government should use direct pricing policy on LPG productrnand regulated and pre detennined kerosene pricing policy on kerosene product tornincrease its revenue. This kerosene price should be set below LPG price. As the result,rnthe government can earn more revenue fiom'both products.