ObjectiveTo provide complete, robust data on annual SLE incidence rates over nearly two decades from the Southeast Norway area (2.9 million inhabitants) and assess accuracy of SLE‐specific ICD codes for SLE diagnosis.MethodsFrom administrative databases, we identified all cases ICD‐10 coded as SLE during 1999‐2017 in Southeast Norway. We manually reviewed the chart of every case ICD‐10 coded as SLE to either confirm or reject SLE diagnosis. Using SLE classification criteria, we classified all cases with confirmed SLE. We estimated annual incidence rates of classified SLE, and subsets, defined by age at diagnosis, sex, and parental country of birth. Chi‐square test was applied for linear time‐trend analyses of incidence.ResultsAmong the 3488 cases ICD‐10 coded as SLE, chart‐reviews confirmed SLE diagnosis in 1558 (45%), of which 797 had new‐onset disease during 1999‐2017. Annual SLE incidence‐rates fell during 1999‐2017. The fall was most pronounced in female subjects aged 50‐59 years at diagnosis where incidence fell from 3.4 to 1.1 per 100 000 (p‐trend <0.001). Concurrent ecological data from the study area showed a 74% reduction in prescriptions of menopausal hormone treatment. Accuracy of ICD‐10 codes for incident SLE diagnosis was acceptable in juveniles and young adults (up to 20 years) but otherwise low.ConclusionIn a presumably complete population‐based cohort, we identified decreasing incidence of SLE, especially among female subjects aged 50‐59 years. While reasons for declining incidence are not clear, ecological data indicate a possible role of environmental factors, for example menopausal hormone treatments.image