This study investigated the usefulness of the Black Box Test of Learning Ability as an indicator of children's performance on math and writing tasks. Twelve second grade students, seven to eight years of age, demonstrated naivete on both tasks and were subsequently individually administered the learning test. The subjects were divided into two groups, and each group received a different task presentation order. Composite scores were derived for all subjects and tasks, and individual learning curves were compared. The combined Black Test score produced a significantly high correlate to math (rho = .733) and writing (rho = .841) than either the paired associate or sequential learning tasks alone. Although the length of the learning curve of the BBT was indicative of the learning curve lengths for each school task, trial by trial learning curve comparisons were less reliable. Analysis of these data suggested that the method of instruction (i.e., the amount of attempted practice and appropriate feedback) was the major factor correlating the Black Box Test to each school task. It was suggested that the most useful assessment of "learning ability" would employ a behavioral sample of the task itself, rather than utilize a correlated activity.