Workplace training aims to enhance the performance of trainees by improving their specific skills. unfortunately, the outcome of such trainings in terms of the percentage of increase in trainees’ performance is not generally measured. on the one hand, this lack of measurement results in contractors not recognizing the importance of training in enhancing workers’ performance, and on the other hand, the training providers do not receive constructive feedback on the efficacy of their sessions. in addition, in the construction industry the competency requirements for mid- and upper-level employees are very well-defined, such clear definitions at the workers’ level are missing. this study therefore attempts to bridge this gap by conducting a nationwide survey of 598 industrial professionals in india, to identify six, nine, and thirteen competencies, through descriptive statistics, for unskilled, semi-skilled, and skilled categories of workers, respectively. these competencies included attitude, motivation, and physical strength for unskilled workers; self-confidence and technical knowledge for semi-skilled workers; and writing, reading, mathematical, and problem-solving competencies for skilled workers. further, the authors have measured the identified competencies of the workers at construction site using survey instruments, which were identified through literature. along with the measurement of the competencies, the productivity of the same workers was also measured via direct observation at sites. by applying various regression models to the collected data (measured competencies and productivity), best-fit curves were developed between each of the competencies and observed productivity. subsequently, a regression model was developed by considering all the measured competencies as explanatory variables and productivity as dependent variable. thereafter, the authors visited four training centers in india and measured the competencies of 137 trainee workers periodically (in every 2 days of a 90-day training). this periodical data on competencies of the workers was fed into the developed regression model to compute their performance increment during their training. subsequently, a learning curve was drawn between the performance increment of the workers and their training time. in addition, the competencies versus training time curves were also plotted to understand how competencies of the workers were enhancing during their training. the results indicate that workers’ performance after training was far below the desired level of employers; the reason for this emerges from competencies versus training time curves, where it was found that training does not focus on enhancing socio-emotional and cognitive competencies. therefore, the study recommends that training providers incorporate these competencies in the training curriculum to achieve more desirable outputs from training. this will help in designing an effective workers’ training program.