Many neurological therapeutic trials require a longitudinal assessment of cognitive functions. An ideal instrument for that purpose should be in accordance to the criteria of classical testing theory and, furthermore, it should be repeatable and economic in administration and interpretation. We developed NeuroCogFX, a computerized assessment battery, according to these criteria. NeuroCogFX comprises subtests for short term memory, working memory, psychomotor speed, selective attention, verbal and figural memory and verbal fluency (mean duration: 25 minutes). Age-related normative data was obtained from 244 subjects without history of neurological or psychiatric disease (age range 16 - 75 years). Forty-two subjects were re-tested after an average of 8 weeks (range: 6 - 10 weeks) in order to assess retest reliability and training effects. Retest-reliabilities were middle-sized in all but one subtest, ranging from r (12) = 0.5 to r (12) = 0.7 (2-back Test: r (12) = 0.37). For construct validation NeuroCogFX was administered in addition to a comprehensive neuropsychological assessment battery in a group of 40 healthy subjects and in 42 patients with chronic epilepsy. The test allows a valid assessment of short-term memory, reaction speed, memory and verbal fluency. NeuroCogFX is an economic, sufficiently reliable and valid instrument for the neuropsychological follow-up examination in single patients and study groups which can be administered if a comprehensive neuropsychological assessment is unavailable.