Asthma is a complex syndrome with multiple phenotypes and endotypes, and current diagnostic tools often fail to address this, therefore there is an increasing need for new predicting biomarkers for asthma which can enable personalized treatment strategies. Biomarkers can facilitate earlier diagnosis of asthma, potentially identifying the disease before clinical symptoms appear. This allows for early intervention, preventing asthma progression and complications and improving long-term outcomes. Reliable biomarkers can help monitor disease activity and treatment efficacy in real time. They can also lead to better disease management, reducing hospitalizations, emergency visits, and overall healthcare costs associated with poorly controlled asthma. Severe asthma patients often do not respond well to standard treatments. New biomarkers can help identify these patients and guide the development of novel therapeutic approaches specifically for severe asthma. In this review we summarize different candidate biomarkers on several levels of investigation: genetic, subclinical inflammatory, allergen-specific immune response, and trained immunity, which could be promising for the early detection of asthma and better management of exacerbation and severity of symptoms in children.