Background: Asymptomatic skin sensitization (AS) is a risk factor for the development of allergic symptoms. A meticulous definition of this condition requires a systematic assessment of clinical symptoms before inclusion. Objective: To examine the concordance between retrospective assessment of seasonal allergic symptoms and prospective seasonal symptom registration among subjects with AS. Methods: On the basis of a population survey, autumn 2002, including skin prick tests (positive if >= 3 mm) and a screening questionnaire, 87 subjects with AS to birch and/or grass pollen, birch and/or grass pollen allergic symptomatic subjects (n = 63) and healthy controls (n = 40) were included in January to March 2003, completed diary cards on symptom and medication use during the relevant seasons 2003, and were examined at follow up in autumn 2003. Allergy: positive SPT and symptoms >= seven diary days. Results: Eleven AS subjects (birch: n = 10) subsequently developed allergic symptoms, yet nine admitted, at follow up, to have had symptoms before inclusion, or even denied pollen-related symptoms despite a significant diary. Compared with AS subjects sensitized to grass pollen, AS subjects sensitized to birch pollen had significantly larger skin prick reactions and more often and severe pollen symptoms. Conclusion: In the context of double-sensitization, retrospective symptom assessment is not a reliable method for ensuring that subjects classified, as asymptomatically skin sensitized, are truly, asymptomatic. This matter should be considered in studies on allergy development.