The purpose of this study was to gather construct validity evidence for a pain classification system for advanced cancer patients using content experts. Two expert panels, representing regional (Panel A, n = 18) and national/international (Panel B, n = 52) palliative medicine and pain specialists, were purposefully selected to participate in a modified Delphi survey technique, to evaluate an existing pain classification system, the Revised Edmonton Staging System (rESS). Each panel participated in two survey rounds, with response rates of 67% (Panel A, Round 1), 39% (Panel A, Round 2), 56% (Panel B, Round 1) and 64% (Panel B, Round 2). The rESS consists of five features: mechanism of pain, incidental pain, psychological distress, addictive behavior and cognitive function. Most participants either agreed or strongly agreed with including the five existing rESS features in a pain classification system, ranging from 67% (Panel A, cognitive function) to 100% (Panel B, mechanism of pain). Most participants suggested keeping the current definitions for these features, with some revisions. Based on participant feedback, definitions for incidental pain, psychological distress, addictive behavior and cognitive function were revised, including the development of guidelines for use. To reflect its intended use as a classification system, the name of the instrument was changed to the Edmonton Classification System for Cancer Pain (ECS-CP).