Abstract: The optimal plans of life tests is investigated under cost constraints when the competing risks data are progressive Type I interval censored with binomial removals. It is assumed that the lifetimes of the individual causes are statistically independent and exponentially distributed with different parameters. The maximum likelihood and the Fisher information matrix are derived. A cost function is set up. For given restricted budgets, the number of test units, the number of inspection and the lengths of inspection intervals are determined to minimize the determinant of asymptotic variance of the estimated lifetime parameters (D-optimality). Furthermore, the sensitivity analysis is conducted through Monte-Carlo simulation to investigate the effects of different lifetime parameters, removal probabilities, as well as cost coefficients on the optimal plans.