: Asymptomatic patients with primary chronic immune thrombocytopenia (ITP) are not recommended treatment if their platelet counts are above 30?×?10/l. Factors such as age and comorbidities may influence clinical manifestations and should be considered for treatment decisions. The aim of this study was to determine the impact of clinical characteristics for initiation of ITP treatment, and the patterns of ITP treatment given. We performed an observational cohort study in Sweden with information from medical records and National Health Registers. Adults diagnosed with incident primary ITP between years 2009 and 2016 were included. Multinomial logistic regression was used to assess the impact of factors predicting treatment start. Out of 858 patients with chronic ITP from 71 hospitals we identified 585 (68%) with a first ITP treatment. For 537 (92%) corticosteroids were the first choice. The median platelet counts at start of treatment was 12?×?10/l (interquartile range 5-27?×?10/l). The variables predicting treatment start were platelet counts below 20?×?10/l and treatment with antihypertensive drugs. Patients with diabetes were less likely to receive corticosteroids. Severe bleeding occurred in 75 (13%) of the patients. Platelet counts below 20?×?10/l, antihypertensive treatment and bleedings were the strongest predictors of treatment start, diabetes yielded lower odds to start corticosteroid treatment. The majority of the patients had corticosteroids as first treatment while second treatment was diverse. Asymptomatic thrombocytopenia is not considered a reason as such for initiating treatment. In the latter years, splenectomy seemed to occur later in the course of treatment.