This study was conducted in order to identify gifted children attending kindergartens of elementary schools, determine how successful families and teachers were in selecting these children, and see how consistent their opinions were with children's actual performance. Participants were children attending kindergartens of elementary schools, their teachers and parents. The identification procedure used in the first stage of this relational survey study involved Parent Observation (POF) and Teacher Observation Forms (TOF) for teachers and/or parents to nominate potentially gifted children, the Primary Mental Abilities Test 5-7 (PMA 5-7) in the second stage and Goodenough-Harris Draw-a-Person Test for children. A total of 113 children out of 600 kindergarteners in central Duzce were nominated by their teachers and/or families, went through the identification procedures, and constituted the sample. This research indicated that teacher and parent opinions had a 44.3% success rate in determining gifted children (50 children). It was found that families were better than teachers in identifying gifted children; teachers made more realistic evaluations of children's performance as shown by tests and scores; but children's actual performance was much better than teacher and family opinions. No meaningful relationship existed between the PMA 5-7 and Goodenough-Harris Test scores of children who were identified as gifted. The Goodenough-Harris Draw-a-Person Test was included in the study to support the results of the PMA 5-7 Group Intelligence Test. The lack of a relationship between scores obtained from these two may be attributed to the facts that Turkish children started preschool education with a delay and were generally given little or no chance by their families to practice activities for thin motor muscle development on their own.