To validate a set of peer nomination forms, 40 different prototypical descriptions were written to represent four aptitude domains (intellectual, creative, socio-affective, physical) and four talent fields (academic, technical, artistic, interpersonal). They were distributed among six experimental forms, which were paired and administered in counterbalanced oreder at a 3-week interval to a sample of 2350 pupils about equally divided between both sexes and three grades (4, 5, 6), and to the teachers of their 88 groups. Indices of reliability —interpeer agreement— were computed for each prototype in every group. A majority of the prototypes maintained acceptable interpeer agreement levels; important variations were observed associated with a small age effect, a significant group effect, and a very large prototype effect, confirming that interpeer agreement depends first and foremost on the specific aptitude or talent assessed. Consensus among peers was best for intellectual and physical aptitudes, and academic and artistic talents; it was worst for socioaffective aptitudes and interpersonal talents.