The influence of different diagnostic approaches on familial aggregation of spelling disability was investigated in three studies. In the first study, in a sample of 32 dyslexic children and their families, we found significantly increased rates of spelling-disabled sibs and parents by applying the IQ-discrepancy criterion. There was no evidence for the assumption that IQ-discrepancy and low achievement criteria define different subgroups of spelling disorder regarding familial aggregation. In the second study, in a sample of 79 adults, it could be demonstrated that questionnaire data can be used as an appropriate method to classify adult probands as spelling disabled with a correct classification rate above 87%. In the third study, a subgroup of dyslexic boys could be characterized by a lack of the N1-component in visual evoked potentials which was most prominent in those boys whose spelling scores were more than 1.5 standard deviations below their intelligence level. This subgroup could be interesting also for genetic research.