This research focuses on the parallel machines scheduling problem in the hard disk drive industry, with the objective of minimizing the makespan. In this problem, unrelated parallel machines with sequence-dependent setup time and machine eligibility are considered. This problem can be formulated as a deterministic R m |M j ,S jk |C max problem. To solve the problem, a mathematical model is developed for small-sized problems. For large-scale problems, algorithms based on a constructive heuristic (Unrelated Parallel Machines with Sequence-Dependent Setup Times and Machine eligibility (UPSDSTM)) and Differential Evolution (DE) are applied. The results demonstrate that the DE algorithm is very efficient. It yields higher relative improvement (RI) on the makespan.