Many research questions require the assessment of reading in large samples of children. We compared two nonconventional procedures for early reading assessment: test administration by telephone, and teacher assessment. Five thousand five hundred and forty four children participating in a longitudinal twin study were assessed by telephone using the Test of Early Word Reading Efficiency, and by their teachers using the UK National Curriculum criteria. A correlation of 0.69 was obtained between the TOWRE and Teacher Assessment for Reading. There was also good agreement between the two procedures for identification of children at the lower extreme. Both appear to be practical and valid for research studies.