Objectives: Fish roe is a common allergen in Japan. We have previously reported that although immunoglobulin (IgE) from patients with salmon roe (SR) or pollock roe (PR) allergies cross-react, 70% of patients with SR allergy can consume PR without developing any symptoms. However, a correlation between clinical cross-reactivity and serological cross-reactivity remains to be demonstrated. Methods: Serum samples were collected from 15 patients with SR allergy who had consumed cooked PR previously. Among these volunteers, four had experienced immediate symptoms after consuming cooked PR, while the others had exhibited no symptoms of PR allergy. A competitive enzyme-linked immunosorbent assay (ELISA) was performed to analyze the serological cross-reactivity with SR and PR. Immunoblotting inhibition assays were performed using serum samples that had been pre-incubated with SR or PR extracts. Results: In ELISAs, binding to SR was inhibited by >50% when the serum samples from patients with both SR and PR allergies were pre-incubated with PR extract (p=0.0256). In immunoblots, pre-incubation of serum samples with PR extract inhibited detection of the 16-kDa protein, which likely corresponds to the major SR allergen beta' component, significantly more for samples from patients with both SR and PR allergies (100%) than for samples from those with only an SR allergy (18.2%) (p=0.011). Conclusions: The superior competitive binding of the sera from patients with both SR and PR allergies to PR compared with that to SR may induce clinical cross-reactivity between SR and PR.