Aurora kinase A, B and C, are key regulators of mitosis and are over expressed in many of the human cancers, making them an ideal drug target for cancer chemotherapy. Currently, over a dozen of Aurora kinase inhibitors are in various phases of clinical development. The majority of the inhibitors (VX-680/MK-0457, PHA-739358, CYC116, SNS-314, AMG 900, AT-9283, SCH-1473759, ABT-348, PF3814735, R-763/AS-703569, KW-2449 and TAK-901) are pan-selective (isoform non-selective) and few are Aurora A (MLN8054, MLN8237, VX-689/MK5108 and ENMD 2076) and Aurora B (AZD1152 and GSK1070916) sub-type selective. Despite the intensive research efforts in the past decade, no Aurora kinase inhibitor has reached the market. Recent evidence suggests that the sub-type selective Aurora kinase A inhibitor could possess advantages over pan-selective Aurora inhibitors, by avoiding Aurora B mediated neutropenia. However, sub-type selective Aurora kinase A inhibitor design is very challenging due to the similarity in the active site among the isoforms. Structural biology and computational aspects pertaining to the design of Aurora kinase inhibitors were analyzed and found that a possible means to develop sub-type selective inhibitor is by targeting Aurora A specific residues (Leu215, Thr217 and Arg220) or Aurora B specific residues (Arg159, Glu161 and Lys164), near the solvent exposed region of the protein. Particularly, a useful strategy for the design of sub-type selective Aurora A inhibitor could be by targeting Thr217 residue as in the case of MLN8054. Further preclinical and clinical studies with the sub-type selective Aurora inhibitors could help bring them to the market for the treatment of cancer.