For design of stiffened composite shells, for example aircraft fuselage, it is important to know actual laminated composite properties. Conventional way is testing the coupon specimens, which are manufactured with similar technology as the real large structure. Employing such method still is open the question, whether the material properties obtained from the coupon tests are the same as in the large structure. In the present paper comparative study of three methods is performed. Material properties of laminated carbon/epoxy composite were obtained from conventional static tests, from vibration tests of the flat plates and from vibration tests of the small panel with one stiffener. The small panel was cut from the large stiffened composite panel, which before was tested for post-buckling. The method of identification is based on solution of inverse problem. For this experimental and numerical frequencies are employed in identification functional, which is defined as squared discrepancies between the measured and numerical frequencies. Approximation of the functional was obtained using method of experimental design and response surface approach [1, 2]. The results for elastic properties obtained employing the first (static tests of coupons) and second (vibration tests of flat plates) method are in good agreement. The results obtained from the vibration tests of the small panel cut from the real structure are with some differences. This can be explained so, that single ply thickness, material density, layer angles of the real structure is slightly different from the nominal values. These differences should be taken into account in design of real structures by choosing safety factors, calculating the limit and the collapse loads of composite stiffened structures.