The Convention on Biological Diversity (CBD) has adopted a scheme of using scientific criteria for identifying 'Ecologically or Biologically Significant Marine Areas' (EBSAs) in need of protection in open ocean and deep-sea habitats. To date, expert opinion collated during regional workshops has been the main method to identify regional EBSAs. In this paper, we propose a new method that could complement this process by adding more objective and transparent analyses. There are four main steps: 1) identify the area to be examined, 2) determine appropriate datasets and thresholds to use in the evaluation, 3) evaluate data for each area/habitat against a set of criteria, and 4) identify and assess candidate EBSAs. The method can be applied to any habitat, but offshore seamounts were chosen as a test habitat to develop and evaluate it. Several options for various combinations of criteria are presented, with one being proposed as the most appropriate to identify a tractable number of seamounts that satisfied the EBSA criteria and which could be combined into larger areas that represent meaningful ecological and practicable management units. This option selects seamounts that meet any one of the 5 "biological" criteria (i.e. unique/rare, diverse, productive, threatened species, critical habitat) and which contain environmental features that are vulnerable to human activities but not yet significantly impacted by them. This selection process resulted in 83 seamounts being identified from over 3000 evaluated in the South Pacific Ocean. The priority seamounts group into 10 areas, consisting of 5 clusters of seamounts, and 5 individual seamounts. The primary strength of the method is the adoption of a transparent, and logically sequential, selection process that is conceptually transferrable to other habitat types and regions beyond our model system. We contend that in a global EBSA context it can be a useful tool to assist deep-sea management. (C) 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).