The phenotype of four cases of Smith-Lemli-Opitz syndrome (SLO) with proven defects in cholesterol biosynthesis are compared, and shown to be markedly disparate even between sibs, and demonstrate the dilemma for the clinician. The advent of a biochemical test for SLO has been enormously valuable in determining which patients are truly affected by the condition, but because of the wide phenotypic variation, a diagnosis on clinical features alone remains problematic.