β-lactoglobulin (BLG) is the primary allergen in bovine milk allergy. The identification of key amino acids in the BLG epitopes has not been comprehensive due to differences in identification methods, patient symptoms, and population characteristics. In this study, bioinformatics predictions were conducted for key amino acids based on two potential IgE linear epitopes in BLG. Then, the peptide AA30-43 was confirmed as an IgE linear epitope through alanine scanning mutagenesis and peptide microarray assays, with four key amino acids (A34, A37, R40, and V41) common to different symptoms being identified. Moreover, symptom-specific key amino acids were identified. Serine (S30) and aspartic acid (D33) are the key amino acids for cutaneous allergy, while food allergy sera showed a preference for recognizing leucines in different positions (L31 and L39). Additionally, mutant peptides (R40, V41, L39, and D33) showed an obvious decrease in digestive stability compared with the epitope. Finally, the results of the KU812 cell degranulation model validated the critical role of the amino acids in allergenicity. These findings offer significant advantages for advancing both immune tolerance therapies and hypoallergenic milk product development, which hold significant implications for further research, prevention, and treatment of bovine milk allergy.