Memory tests were often developed for healthy populations. The accuracy of these measures is reduced when administered to patients with neurological diseases, who may experience physical and/or cognitive symptoms. Also, methodological factors, for example, spanning the ability spectrum, and content/format artefacts, may contribute to a decline in test precision. The aim of this study was to develop a new test of memory, which addresses these issues. The new memory test comprises assessments of recall, paired association, and recognition, at a Task Familiarisation stage and two difficulty levels, for both the verbal and spatial modalities. It was administered to 85 healthy individuals and 100 patients with multiple sclerosis (MS). All patients were able to attempt each task of the new assessment and there was no influence of visual integrity or manual dexterity on memory test performance, supporting the applicability of the tasks to patients with multiple sclerosis. Both the standardisation and validation samples demonstrated a wide range of scores on each section of the new test suggesting that the measure spanned an acceptably broad range of abilities. It seems probable, therefore, that the new assessment offers a more exact measure of verbal and spatial recall, paired association, and recognition memory.