The role of non-invasive methods in the investigation of acute effects of traffic-related air pollution is not clearly established. We evaluated the usefulness of non-invasive biomarkers in detecting acute air pollution effects according to the age of participants, the disease status, their sensitivity compared with lung function tests and their specificity for a type of pollutant. Search terms lead to 535 titles, among them 128 had potentially relevant abstracts. Sixtynine full papers were reviewed, while 59 articles were excluded as they did not meet the selection criteria. Methods used to assess short-term effects of air pollution included analysis of nasal lavage (NAL) for the upper airways, and induced sputum (IS), exhaled breath condensate (EBC) and exhaled nitric oxide (FeNO) for central and lower airways. There is strong evidence that FeNO evaluation is useful independently from subject age, while IS analysis is suitable almost for adults. Biomarker changes are generally observed upon pollutant exposure irrespective of the disease status of the participants. None of the biomarkers identified are specific for a type of pollutant exposure. Based on experimental exposure studies, there is moderate evidence that IS analysis is more sensitive than lung function tests, whereas this is not the case for biomarkers obtained by NAL or EBC. Cells and some cytokines (IL-6, IL-8 and myeloperoxidase) have been measured both in the upper respiratory tract (NAL) and in the lower airways (IS). Overall, the response to traffic exposure seems different in the two compartments. In conclusion, this survey of current literature displays the complexity of this research field, highlights the significance of short-term studies on traffic pollution and gives important tips when planning studies to detect acute respiratory effects of air pollution in a non-invasive way.