Career-related continual learning often involves formal and informal, off- and on-the-job learning, as characterized by the three types: formal off-the-job classroom learning, informal self-training, and task characteristics learning. A questionnaire survey and Structural Equation Modeling serves to test propositions pertaining to the interactive effects of these three categories of learning. The model derives from implicit belief and self-efficacy theories, which stress how personal beliefs help translate external data. A career stage and work characteristic model also provides guidelines for designing managerial task characteristics for engineering. The sample, representing 449 engineering careerists from six technical enterprises in Taiwan, comprises both R&D and general engineers and managers. The results suggest that the best managerial learning occurs after engineers practice management at work. Of the three learning methods, self-training engenders the most benefits, though the creation of these benefits depends on the task characteristics of the on-the-job learning opportunities offered by organizations. Namely, the combined way generates the best learning way for management in engineering. Off-the-job classroom training should be provided to young engineers, who have not yet had chances to learn management in the workplace. Overall, the findings suggest that management is learnable; however, it requires proper multiscale measurements to reflect engineers' actual learning experience.