The presence of imperfections significantly reduces the load-carrying capacity of thin cylindrical shells and the reduction depends on the shape and the size of the imperfections. As a result, the prediction of the shells' buckling capacity requires a priori knowledge about the imperfections---a difficult, expansive, and time-consuming adventure, if not impossible. Consequently, thin cylindrical shells are designed conservatively using the knockdown factor approach that accommodates the uncertainties associated with the imperfections present in the cylinders; almost all the design codes follow this approach explicitly or implicitly. However, cylindrical shells can be designed more efficiently by making them insensitive to imperfections, or by knowing their exact capacity without the difficult task of measuring the imperfections. This dissertation examines these two approaches for the efficient designing of thin cylindrical shells. In addition, we investigate buckling behavior and imperfection sensitivity of thin cylindrical shells under pure bending along with their applications in tall wind turbine towers. For making thin cylindrical shells insensitive to imperfection, wavy cross-sections are proposed instead of circular cross-sections. Past studies have demonstrated the effectiveness of wavy cylinders to reduce imperfection sensitivity under axial compression assuming linear elastic material behavior and using eigenmode imperfections. In this dissertation, using a realistic dimple-like imperfection, new insights are presented into the response of wavy cylinders under uniform axial compression and bending. We found that thin cylindrical shells can be made imperfection insensitive by manipulating their cross-section geometry. For high-fidelity estimates of the capacity of thin cylindrical shells without measuring the imperfections, a novel procedure is proposed based on the probing of the axially loaded cylinders. Computational implementation of the proposed procedure yields accurate results when the probing is near the imperfection; however, the procedure over-predicts the capacity when the probing is away from the imperfection. It demonstrates the crucial role played by the probing location and shows that the prediction of imperfect cylinders is indeed possible if the probing is at the proper location. The behavior of cylindrical shells under bending and their imperfection sensitivity have not been fully understood for all the range of dimensions. In this dissertation, we investigate the buckling behavior and imperfection sensitivity of thin steel cylindrical shells under pure bending, focusing on a specific range of slenderness, which is found in energy structures such as tall wind turbine towers (60 < R/t < 120). We found that strain-hardening models play an impactful role on the bending behavior; moreover, the presence of imperfections reduces the collapse curvature more than the reduction in peak moment. Further, we propose wavy wind turbine towers to make wind turbine towers efficient. The imperfection sensitivity of the wavy towers is evaluated, and we found that the sensitivity of the wavy towers is small compared to that of the circular towers.