Departing from the observation that illiterates significantly underscore in some neuropsychological tests, a learning-to-read method named NEUROALFA was developed. NEUROALFA is directed to reinforce these underscored abilities during the learning-to-read process. It was administered to a sample of 21 adult illiterates in Colima (Mexico). Results were compared with 2 control groups using more traditional procedures in learning to read. The NEUROPSI neuropsychological test battery was administered to all the participants before and after completing the learning-to-read training program. All 3 groups presented some improvement in the test scores. Gains, however, were significantly higher in the experimental group in Orientation in Time, Digits Backward, Visual Detection, Verbal Memory, Copy of a Semi-Complex Figure, Language Comprehension, Phonological Verbal Fluency, Similarities, Calculation Abilities, Sequences, and all the recall subtests, excluding Recognition. Performance in standard reading tests was also significantly higher in the experimental group. Correlations between pretest NEUROPSI scores and reading ability were low. However, correlations between posttest NEUROPSI scores and reading scores were higher and significant for several subtests. Results are interpreting as supporting the assumption that reinforcement of those abilities in which illiterates significantly underscore results in a significant improvement in neuropsychological test scores and strongly facilitates the learning-to-read process. The NEUROALFA method of teaching reading to adult illiterates is beginning to be used extensively in Mexico. To our knowledge, this is the first attempt to apply neuropsychological principles to social problems. (JINS, 2000, 6, 789–801.)