We sought to assess the interobserver variation of the new International Society of Nephrology/Renal Pathology Society (ISN/RPS) classification of lupus nephritis when compared with the previous World Health Organization classification, when used by pathologists throughout the UK. We also sought differences in how the 2 classifications were applied to a single set of biopsies. Twenty unselected renal biopsies showing lupus nephritis were circulated to pathologists in the UK National Renal Pathology External Quality Assessment Scheme, before the ISN/RPS scheme was published, with a request to apply the World Health Organization classification. The same slides were recirculated approximately 1 year later with a request to apply the ISN/RPS classification. A significant improvement in interobserver reproducibility was demonstrated by the new classification (kappa 0.53 vs. 0.44, P = 0.002). The reproducibility of the assessment of disease activity and chronicity remains suboptimal (kappa = 0.33). The new classification tends to produce more diagnoses of Class IV lupus nephritis, with fewer diagnoses of Classes III and V. The improvement in interobserver reproducibility indicates that an important aim of the new classification has been achieved. Further work is needed to determine whether the increase in diagnosis of Class IV nephritis represents an improvement in biopsy interpretation or a divergence from the previous classification, as the latter could undermine attempts to relate results from the new system to treatment strategies based on clinical trials which used the old.