This paper reviews recent progress in perfect absorbers of electromagnetic waves that utilize metamaterials (MMs). Various numerical techniques that are inevitable for the research on MMs, along with their pros and cons, are commented on and discussed. Abandoning the conventional periodic-boundary conditions is crucial for understanding the discrepancy between simulation and experiments for certain MMs. Broadband perfect absorption can be realized by using multiple resonances and/or multiple units with different sizes in a unit cell, or by utilizing bad Q-factors by incorporating the electrically resistive components, in addition to the ordinary metallic elements. The classical analogue of electromagnetically-induced transparency can be utilized to realize perfect absorption, by employing either structural asymmetry, or near-field interactions. Perspectives and suggestions for further study are discussed.