The biases in volutary participation by adult twins are well known but less attention has been paid to twin children where parents decide on participation and provide much of the information. Several aspects of reporting including the assessment of zygosity are compared in four large Australian data bases: 1) a nationwide compulsory (and hence representative) survey of literacy and numeracy; 2) a nationwide “Twins in School” survey of parents and teachers of twins run through Education Departments and AMBA, the parents organisation in conjunction with LaTrobe; 3) the LaTrobe Twin Study which is a longitudinal program involving frequent interactions between families and researchers, and 4) the Australian NHMRC Twin Registry which has surveyed a large sample of their families with twin children by mail. One potential bias comes when recruitment is on a continuing basis as in the LaTrobe Twin Study and the Australian Twin Registry when differences between “early” and “late” enrolling families arise. One difference between the four samples arose from parents being much more likely to contrast their twins and to report problems in one but not the other, whereas teachers' and psychologists' assessments of these same children generally reported much smaller intrapair differences. Future studies should have some common questions to provide comparative data on such biases. Key questions are proposed for this area, mainly on the perceived need for different forms of remediation, together with other recommendations about the minimal essential baseline data set for a registry.