A study of possible undue neurobehavioral effects of low-level lead exposure was performed in Danish school children who entered first grade in 1982. Lead absorption was found to relate to impaired psychological test performance and difficulties at school. However, this relationship was confounded by medical risk factors for neurological deficit. Such factors were used as exclusionary criteria before psychological testing and were selected a prior on the basis of a critical evaluation of known etiologies of neurobehavioral dysfunction. Children characterized by a medical risk factor tended to show performance below average and, at the same time, low lead absorption. The medical risks were also associated with delayed motor activity in the first year of life, thus perhaps resulting in diminished lead intake. This confounding effect makes strict exclusion for proven medical risk factors crucial to avoid bias toward the null-hypothesis. In spite of strict exclusion, a residual confounding due to health-related variables still remained. Inconsistent findings in the field of neurobehavioral studies may be partly explained by different means of identification and statistical treatment of medical risk factors.