This article investigates high performance work practices used in project-based globally distributed software development firms in Sri Lanka and their effects on job performance. A random sample of 220 employees engaged full-time in globally distributed software development firms in Sri Lanka responded. For the data analysis, descriptive statistics, factor analysis, correlation analysis, and multiple regression analysis were used. Three main high performance work practices were identified by the analysis, namely, performance evaluation, learning and development, and involvement in decision making. These three practices significantly and positively predict job performance of employees attached to project-based globally distributed software development firms.