Three studies are reported on the development of a four-disc version of the Tower of London test of planning ability. The first (n = 138) involved the selection of items based on rational and empirical criteria to provide a short test of graded difficulty suitable for use with children and clinical populations. The second study (n = 480) checked the properties of the 10-item test on a new sample and in addition examined the internal consistency and factor structure of the test. The third study (n = 61) examined the test–retest reliability of the test over a period of 1 month. The difficulty level of the test remained relatively stable from sample to sample and was sensitive to linear trend in performance from age 5 years up to 30 years. Total score did not reflect the action of a single underlying construct but rather appeared to index a number of factors. Scores were reasonably stable over the 1-month period studied, at least for the children's sample employed. The four-disc version is a promising method of assessing planning in children and adolescents in clinical situations.