Because traditional intelligence tests do not allow a reliable and valid estimate of the general cognitive abilities of individuals with IQs lower than 55, a dynamic test of analogical reasoning (Analogical Reasoning Learning Test—ARLT) was constructed for use with this population. The aim of the test is to distinguish individuals who could undergo more ambitious schooling and cognitive training from those who would profit little or none from such an intervention. Internal consistency and test–retest reliability of this measure are high. Discriminant and predictive validity, assessed with another learning test and with a one-month training on inductive reasoning, respectively, are satisfactory. The test is seen as a useful tool for diagnostic differentiation among low performers.