Correlations are important to estimate engineering properties of soils particularly for project where there is a financial limitation, lack of test equipment or limited time. Although correlations are commonly used in the preliminary stage of any project, it is reasonable to assign a unique strength to soils for their respective physical properties. In this thesis, attempts have been made to obtain valid correlations between unconf ined compressive strength with Atterberg limit , dry density and grain size distribution of sub grade soils. In this correlation, twenty two different disturbed and undisturbed soils samples were collected from Sebetamit - Tis Abay road subgrade soils. Various laboratory tests including Atterberg limit , dry density, grain size Analysis, and unconfined compression strength tests were performed on the samples. Grain size analysis shows the soil is silty clay and according to unified soil classification system the soil sample falls in CH (Inorganic clay), MH (Inorganic silt) and in AASHTO they are in a region A-7-5 and A-7-6 (clayey). In this research, statistical software (SPSS) is employed to investigate the significance of individual independent variables. The correlation is established in the form of an equation of unconfined compressive strength as a function of index properties test by considering the effect of an individual soil properties and effect of a combination of soil properties on the unconfined compressive strength value. The developed correlation using single regression analysis the coefficient of determination R 2 = 0.945, while multiple regression analysis generated relatively an improved correlation of R 2 = 0.960, for a sample size of eighteen. After validating the developed correlation with control test results, it was noted that the correlation of unconfined compressive strength value with soil index properties is more applicable for preliminary characterizing the strength of subgrade soils Keywords: Unconfined compressive strength, Index Properties, Prediction Models, Coefficient of determination