Abstract Background and Aims Membranous nephropathy (MN) is a frequent cause of nephrotic syndrome of non-diabetic origin in adults, and a leading cause of end stage renal disease. Its origin is primary in up to 80% of cases but can also be secondary to underlying diseases such as cancer, lupus or other systemic disease, infections or drugs. Important recent advances have been made in our understanding of primary MN physiopathology with the identification of various antibodies such as PLA2R, THSD7A, Exostosin 1,2 suggesting autoimmunity. However, distinction between primary and secondary MN remains difficult and can lead to numerous and expensive complementary exams and postpone treatment. Some authors (Cambier, Ronco, CJASN 2012) proposed an algorithm based on histological and biological features in order to distinguish primary from secondary MN. Yet this algorithm has never been tested. The main objective of our study was to evaluate the efficiency of a combined histological and biological criterion in order to establish the diagnosis of primary MN. Method We conducted a pluricentric retrospective cohort study in Northern France. All patients with histological proved membranous nephropathy between January 2010 and December 2016 were included. All kidney biopsies were re-analysed in order to specify the predominant IgG subclass and PLA2R staining. The combined criterion was considered present if: PLA2R was detected (in biopsy or serum), with no endocapillary proliferation and an IgG4 subclass predominance (or co-predominance) on kidney biopsy. The gold standard to diagnose primary MN was the exclusion of other causes of MN after biological and radiological investigation, in particular the absence of cancer detected after at least 2 years of following. We tested the sensitivity, specificity, and predictive values of this combined criterion in order to diagnose primary MN. Results 173 patients were included. 66 (38.2%) patients were women. At the time of diagnosis mean age was 52.1 years old (SD: 17.4), mean albumine and creatinine were respectively 25.2 g/L (9.02) and 14 mg/L (17.7). Mean protein to creatinine ratio was 5.75 g/g. 119 (68.8%) patients had primary membranous nephropathy. Secondary membranous nephropathy were due to lupus, other auto immune disease and cancer in respectively 25 (14.5%), 14 (8.0%), and 15 (8.0%) cases. 59 patients had a positive combined criterion. Sensitivity of the criterion was: 0.50 95%CI: 0.40-0.59, Specificity was: 0.90 95%CI: 0.80-0.98 compared to a specificity of 0.50 95%CI 0.34-0.54 for single criterion PLa2R on biopsy. Negative predictive value was 0.45 95%CI: 0.31-0.59 and positive predictive value was 0.92 95%CI: 0.84-0.99. Histological features associated with primary membranous nephropathy were: no extracapillary proliferation (p=0.003), presence of PLA2R (p<0.001), IgG4 predominant staining (p<0.001) and low Ig1 staining (2.24 + in secondary MN versus 1.86 + in MN) p=0.011. Conclusion Our histological combined criterion had a high specificity: 0.90 95%CI 0.80-0.98 and positive predictive value 0.92 95%CI: 0.84-0.99. Presence of this combined criterion could allow the clinician to reduce the number of complementary analyses and help conclude faster to the primary etiology of the MN.