<![CDATA[The purpose of this study was to identify and analyze the direct and indirect effects of work demands and self-efficacy variables on job satisfaction through work involvement as an intervening variable for employees of PT. Metromatika Teknologi Rekayasa Medan. The analysis used is associative analysis, this research method is quantitative. The population in this study are all employees at PT. Metromatika Teknologi Rekayasa Medan totaling 70 people. Sampling method using nonprobability sampling method with census sampling technique. So, the sample in this study is 70 employees at PT. Metromatika Teknologi Rekayasa Medan The results obtained from this study indicate that: 1) Work demands directly have a significant effect on employee work involvement at PT. Metromatika Teknologi Rekayasa Medan, 2) Self-efficacy directly has a significant effect on employee work involvement at PT. Metromatika Teknologi Rekayasa Medan. 3) Work involvement directly has a significant effect on employee job satisfaction at PT. Metromatika Teknologi Rekayasa Medan, 4) Work demands do not directly influence significantly the job satisfaction of employees PT. Metromatika Teknologi Rekayasa Medan, 5) Self-efficacy has a significant direct effect on employee job satisfaction at PT. Metromatika Teknologi Rekayasa Medan, 6) Job demands indirectly have a significant effect on job satisfaction through the involvement of employees at PT. Metromatika Teknologi Rekayasa Medan, 7) Self-efficacy indirectly has a significant effect on job satisfaction through employee involvement at PT. Metromatika Teknologi Rekayasa Medan. In substructure I, the Adjusted R square value of 0.617 can be called the coefficient of determination, this means that 0.617 (61.7%) work involvement can be obtained and explained by work demands and self-efficacy while the remaining 38.3% (100% - 61.7 % = 38.3%) explained by variables outside the model that were not examined. In substructure II, the Adjusted R square value of 0.797 can be called the coefficient of determination, this means that 0.797 (79.7%) work involvement can be obtained and explained by work demands and self-efficacy while the remaining 20.3% (100% - 79.7 % = 20.3%) explained by variables outside the model that were not examined.]]>
