This phase I trial examined the safety, pharmacokinetics, and pharmacodynamics of MLN8054, an oral, selective, small-molecule inhibitor of Aurora A kinase. Patients with advanced solid tumors received increasing doses of MLN8054 in 28-day cycles until dose-limiting toxicity (DLT) was seen in ≥2 of 3-6 patients in a cohort. For the 10-mg and 20-mg cohorts, treatment was administered once daily on days 1 to 5 and 8 to 12. Patients in later cohorts (25, 35, 45, 55, 60, 70, and 80 mg/day) were treated four times daily on days 1 to 14, with the largest dose at bedtime (QID-14D) to mitigate benzodiazepine-like effects possibly associated with peak plasma concentrations. Patients (n = 43) received a median of 1 cycle (range, 1-10). DLT of somnolence was first noted in the 20-mg cohort. Two DLTs of somnolence (n = 1) and transaminitis (n = 1) were seen at QID-14D 80 mg. Grade 2 oral mucositis (n = 1), predicted to be a mechanistic effect, was observed only at QID-14D 80 mg. MLN8054 exposure levels were roughly linear with dose; terminal half-life was 30 to 40 hours. Pharmacodynamic analyses of skin and tumor mitotic indices, mitotic cell chromosome alignment, and spindle bipolarity provided evidence of Aurora A inhibition. MLN8054 dosing for 10 to 14 days in 28-day cycles was feasible. Somnolence and transaminitis were DLTs. Pharmacodynamic analyses in mitotic cells of both skin and tumor provided proof of mechanism for Aurora A kinase inhibition. A more potent, selective, second-generation Aurora A kinase inhibitor, MLN8237, is in clinical development.