OBJECTIVE: To estimate the agreement between observers on the detection of goitre by physical exploration in a school population. METHODOLOGY: We performed a cross-sectional study to detect goitre in a representative sample of 1134 schoolchildren aged 6 to 12 years from 20 schools in the health area of La Mancha Centro. The examination was performed blinded by two observers. Five grades in thyroid size were established (0, Ia, Ib, II and III). Above grade Ia was considered as goitre. The agreement was assessed in relation to variables such as age, sex, body mass index, height, and day of examination. The weighted kappa was used to measure the agreement. RESULTS: In the 1097 schoolchildren with a dual examination, 96 (8.8%) cases of goitre were detected by observer 1, and 102 (9.3%) cases by observer 2, (P=.58). The degree of interobserver agreement in the identification and grading of goitre was moderate (kappa 0,55, 95%CI: 0,46 to 0,64) for the first, and substantial (weighted kappa 0,61; 95%CI: 0,51 to 0,71) for the second. The degree of agreement was somewhat higher in girls, older schoolchildren, increased weight, height, and body mass index. The interobserver agreement was relatively stable throughout the study. CONCLUSIONS: The interobserver agreement in detecting goitre by palpation in our study is moderate, but is lower in younger children and stable for the duration of study.