Evaluation of the therapy of X-linked adrenoleukodystrophy (X-ALD) is hampered by its rarity and by the striking and unpredictable variation in phenotypic expression. We present two approaches that may facilitate therapy evaluation. (1) We have analysed data on 377 X-ALD patients who have been followed at the Kennedy Krieger Institute for a mean period of 38 months and have subdivided them into 18 subgroups on the basis of age and the degree of abnormality in brain magnetic resonance imaging (MRI) as assessed by the Loes score (Am. J. Neuroradrol 1994; 15: 1761). We find that grouping on the basis of age and MRI score provides information that is of significant prognostic value. (2) We present plans for the development of a placebo-controlled multicentre international study that will have sufficient biostatistical power to provide objective evaluation of new therapeutic interventions.