Evidence is presented that many primary-age children are good at solving word problems. The data comes from three years of observation of about 100 pupils from grades one through three. It is noted that instances of total failure were relatively rare. The individual interview method is recommended to teachers. (MP)