In the US, the Medicare Modernization Act of 2003 required that Medicare Part D insurers provide medication therapy management (MTM) services (MTMS) to selected beneficiaries, with the goals of providing education, improving adherence, or detecting adverse drug events and medication misuse. These broad goals and variety in MTM programmes available make assessment of these programmes difficult. The objectives of this article are to review the definitions of MTMS proposed by various stakeholders, and to summarize and evaluate the outcomes of MTMS consistent with those that may be offered in Medicare Part D or reimbursed by State Medicaid programmes. MTM programmes are approved by the Centers for Medicare and Medicaid Services (CMS). Pharmacy, medical and insurance organizations have provided guidelines and definitions for MTM programmes, distinguishing them from other types of community pharmacy activities. MTM has been distinguished from disease state management because of the focus on medications and multiple conditions. It differs from patient counselling because it is delivered independent of dispensing and involves collaboration with patients and providers. There is no consensus on the recommended mode of delivery (i.e. face-to-face or by telephone) for MTM. A MEDLINE search was conducted to identify articles published after 2000 using the search terms 'medication therapy management' and 'medication management'. Studies with outcomes evaluating community-based programmes consistent with MTMS, regardless of MTMS reimbursement source, were included in the review. Seven publications describing four MTMS were identified. For each of the identified articles, we describe the study design, service setting, inclusion criteria and outcomes. An additional three surveys describing multiple MTMS were identified and are summarized. Finally, ongoing efforts by CMS to evaluate the success of MTMS in the US are described. To date, there are limited outcomes available for MTMS. The wide variety of MTMS offered means that evaluations of individual programmes are necessary. Despite the potential benefits of MTMS, there are numerous challenges to providing MTMS, including reimbursement, justification of the service and stakeholder acceptance of the services. There remains a need for adequately funded, prospective, controlled studies of MTM programmes using strong designs to advance our knowledge of the effectiveness of various interventions and methods of delivery, and for naturalistic studies assessing the impact of CMS approved MTM programmes.