For a thin-walled beam whose cross section size is very large relative to thickness, the actual stiffness and strength tend to deviate from their theoretical predictions based on classical beam theories. Experimental results for square thin-walled beams indicate that such deviations become appreciable when the size/thickness ratio exceeds 50. Empirical formulas are derived to describe these deviations in terms of size/thickness ratio. This information, if incorporated in a structural analysis procedure, should enhance the fidelity of analysis. The efficiency of thin-walled beams with respect to the stiffness-governed and the strength-governed designs is discussed. A possibility of achieving weight savings through substituting aluminum for steel in a stiffness-governed design is also discussed.