There has been a dramatic increase in recent years in the number of papers, reports, etc.,which have been published concerning Marine Protected Areas (MPAs). This overview of the objectives, selection, design and management of MPAs aims to provide a basis for discussion regarding possible ways forward by identifying emerging issues, convergences and divergences. Whilst the attributes of the marine environment may limit the effectiveness of site-specific initiatives such as MPAs, it is argued that it would be defeatist in the extreme to abandon MPAs in the face of these limitations. Ten key objectives for MPAs are discussed, including that of harvest refugia, and it is argued that whilst these objectives may be justifiable from a preservationist perspective, they may be objected to from a resource exploitation perspective. MPAs generate both internal (between uses) and basic (between use and conservation) conflicts, and it is argued that these conflicts may be exacerbated when scientific arguments for MPAs are motivated by preservationist concerns. It is reported that a minority of MPAs are achieving their management objectives, and that for the majority insufficient information was available for such effectiveness evaluations. Structure and process-oriented perspectives on marine conservation are discussed. It is argued that there are two divergent stances concerning optimal MPA management approaches: top-down,characterized as being government-led and science-based, with a greater emphasis onset-aside; and bottom-up, characterized as being community-based and science-guided, with a greater emphasis on multiple-use. Given the divergent values of different stakeholders, the high degree of scientific uncertainty, and the high marine resource management decision stakes, it is concluded that a key challenge is to adopt a "middle-ground'' approach which combines top-down and bottom-up approaches, and which is consistent with the post-normal scientific approach.