In metals, plastic strain localization occurs not only within a broad range of materials but also under different loading conditions and applications. The mechanism of plastic strain localization plays an important role in the processes of plastic deformation of materials and thus the suitability of a given material for a specific structural application. Previous studies have mainly dealt with three general forms of strain localization that develop in metals including Lüders band formation, Portevin-Le-Chatelier (PLC) effect, and necking at the macroscopic scale. Nowadays, the progress and application of elastoplasticity is significantly promoted in industries (e.g., construction, automotive, aerospace). Therefore, with regard to necking, the research to date has tended to focus on the static elastoplastic characterization of metals by means of experimental tests, theoretical modeling and numerical simulations to determine the equivalent stress and strain on the necking section of a tensile specimen. However, accurate estimation of stress and strain distribution at the neck band of a tensile specimen capable of undergoing large postnecking deformation remains unclear due to its complexity. Besides, up to now, far too little attention has been paid to localization length and its evolution throughout loading history, which is essential to predict the postpeak behavior of metallic materials. The overall objective of the research reported in this dissertation is to expand the knowledge of strain localization and ductility of structural metals. At the first stage of this research, strain localization is investigated experimentally in three types of metallic materials, including mild steel Q235, high strength low alloy steel HSLA350, and aluminum alloy AL6061. Displacements, strain distributions, and instantaneous cross-sectional areas of the specimens were measured using a three-dimensional digital image correlation (3D-DIC) measurement system. True stress-strain relations of the specimens were obtained. A new method is proposed to determine the necking zone length from both the longitudinal and transverse strain profiles. At different locations within the necking zone, true stress is distributed unevenly. It is found that the necking zone length increases with the increasing axial elongation of the specimens. At the second stage, a new theoretical method is developed to predict the elastoplastic behavior of the tested materials. Using this method, only basic experimental parameters are needed for estimating the pre-necking true stress-strain relation. The evolution of the neck profile of a cylindrical tensile specimen is derived based on experimental measurements. The pre-necking model is able to describe the Lüders deformation phenomenon with great precision. It can also provide precise predictions for the pre-necking behavior of structural steels S275, S355, S460, and S690. In addition, the new post-necking model can accurately determine the equivalent stress-strain relation. The complicated analysis of the stress correction factor is avoided in this method. Comparison between the new method and the existing ones is carried out. The accuracy of the new model prediction is higher than existing methods, e.g., the Hollomon method, the Ludwick method, and the weighted average method (WAM). At the third stage, two new theoretical models are proposed to respectively predict the strain hardening exponent and the fracture strain, which are important properties of metallic materials. The effect of necking zone length on postpeak stress and strain is also examined. The strain hardening model is developed based on the Swift hardening law. In the case of monotonic loading, the new ductile fracture model is developed independently of stress triaxiality and Lode angle parameter. The strain hardening exponents predicted by the new model are in good agreement with experimental results. The new method provides significantly better predictions in comparison to nine existing models. In addition, the new fracture model provides more accurate predictions than seven popular ductile fracture criteria, and it only requires the initial cross-sectional area and tensile stress of a specimen. Furthermore, it is found that beyond the peak load of a tensile metallic specimen, the stress and strain highly depend on the necking zone length. This research provides an in-depth understanding of the evolution of necking in metals, which is of significant importance to study the post-necking and failure behavior of metallic materials as well as for calculation of ductility of metal structures. In addition, this research not only provides new insights into the elastoplasticity and fracture of metals, but also holds promise in the design of metal forming. Furthermore, the findings of this research can be used in the selection process of metallic materials for diverse applications in construction, automotive, and aerospace industries.