SummaryBackground Allergy and autoimmunity are two potential outcomes of a dysregulated immune system, but the relationship between them is unclear. It has been hypothesized that they could be inversely associated because of different T helper cell reactivity patterns. However, both positive and negative associations have been reported. Therefore, our aim was to perform a large epidemiological study with a defined allergic disease cohort.Methods During the years 1990–2002, 68 770 subjects were tested for total serum IgE (Total‐IgE) and 72 228 were tested with Phadiatop for diagnosing allergic disease at Karolinska University Hospital, Stockholm, Sweden. This cohort was then linked with the Swedish Inpatient Registry 1968–2004 for a follow‐up with regard to recorded discharges for 28 autoimmune diseases. We then used Cox regression and logistic regression to estimate the risk of autoimmune diseases in general in the allergy‐tested subjects.Results Subjects with positive Phadiatop test were at a statistically decreased risk of subsequent autoimmune disease in comparison with subjects with negative test; hazard ratio (HR): 0.80 [95% (Confidence interval) CI: 0.68–0.94). Prior autoimmune disease was associated with a decreased risk of positive Phadiatop test [odds ratio: 0.83 (95% CI: 0.72–0.96)] in comparison with negative test. Subjects with highly elevated Total‐IgE were at a statistically increased risk of a subsequent autoimmune disease in comparison with subjects with normal levels [HR: 1.36 (95% CI: 1.09–1.70)], but no association was found between prior autoimmune disease and different Total‐IgE levels.Conclusion The study supports the hypothesis that allergy, defined as positive Phadiatop test, could be inversely related to autoimmune disease but this association is weak.