Objectives To develop and validate a quantitative radiographic scoring system, the Ghent University Scoring System (GUSS), with better ability to detect progression over a shorter period of time in erosive osteoarthritis (OA) of the interphalangeal (IP) finger joints compared with the existing anatomic phase scoring system. Methods Thirty IP finger joints showing erosive features at baseline or follow-up were selected from 18 patients with erosive hand OA. Posteroanterior radiographs of these joints obtained at baseline, 6 and 12 months-totalling 90 images-were used for the study. All joints were first scored according to the original anatomic phase scoring system. Erosive progression and signs of repair or remodelling were then scored by indicating the proportion of normal subchondral bone, subchondral plate and joint space on an 11-point rating scale (range 0-100 with 10 unit increases). Inter-and intrareader reproducibility was studied using intraclass correlation coefficients (ICCs). Based on the within-variance of two readers, the smallest detectable change (SDC) was calculated and allowed identification of joints with changes above the SDC as 'progressors'. Results Longitudinal inter-reader ICC scores rated well for all variables and the total score (ICC 0.86-0.93). To identify 'real' change over background noise, a change of at least 40 units on the total score (range 0-300) over 12 months (SDC 0-12: 36.0), and 50 units over 6 months (SDC 0-6: 47.6) had to be present. 60% of the 30 joints were identified as 'progressors' over 6 months compared with 33.3% with the classical anatomical scoring system, and 70% versus 56.6%, respectively, over 12 months. Conclusion GUSS, is a reliable method to score radiographic change over time in erosive IP OA and detects more progression over a shorter period of time than the classical scoring system.