Essentials Infections may trigger autoimmune disease and be a complication of an impaired immune system. The cohort, diagnoses and treatments were retrieved from the Swedish Health Registers. Compared to the population, infection risk is increased before primary immune thrombocytopenia. The incidence of cITP (new definition, International ITP Working Group) is 2.30 per 100 000 person‐years. Summary Background Infections after diagnosis of primary chronic immune thrombocytopenia (cITP) have mostly been connected to the immunomodulation treatment. Infections may trigger autoimmune diseases and may be a complication of an already impaired immune system. Objectives To investigate the association of cITP with infection before diagnosis. We also estimated the incidence of cITP based on the new definition by the International ITP Working Group. Methods We identified 1087 adults with primary cITP between 2006 and 2012 using the Swedish Patient Register. Data on infections not already associated with secondary ITP were also retrieved from the register. The standardized incidence ratios (SIRs), using the rates from the general population, and 95% confidence intervals (CIs) were estimated as a measure of relative risk. We used data from the Prescribed Drug Register to estimate SIR for anti‐infective treatment. Results The incidence of cITP was 2.30 per 100 000 person‐years (95% CI, 2.15–2.45). cITP was associated with an increased risk of serious infections requiring inpatient or outpatient care within 5 years before cITP diagnosis (SIR = 8.74; 95% CI, 7.47–10.18). Higher magnitude SIRs were observed for candidiasis, viral infection at an unspecified site and acute upper respiratory infections. For anti‐infective drugs the SIR was 1.37 (1.25–1.50) and the highest SIRs were observed for amoxicillin, macrolides, nitrofurantoin and antivirals. Conclusion Patients with cITP have increased risks of infection and anti‐infective treatments before their cITP diagnosis, with a more marked risk for candidiasis and viral infections. The findings indicate that infection is not only related to the immunomodulation treatment but also to the disease itself.