This manuscript presents a study in which the new version of the computer-based training programme, The Number Race (NR), was used as an intervention for mathematically low-performing children in grade one (M[subscript age] = 86.46, SD = 3.89). In addition to ordinary teacher instruction in mathematics, the intervention group (n = 29) received NR training for 15-minute sessions, 3-4 days per week, during a four-week period. One comparison group comprising mathematically low-performing children (n = 27) and another comprising average-performing children (n = 278) received only ordinary teacher instruction in mathematics during this period. The children's mathematical skills (e.g. counting and basic arithmetic skills) were measured three times during grade one, using three parallel tests. The grouping was based on the first assessment, using the lowest 20th percentile as the cut-off point in the test. The NR intervention took place between the second and third assessments. There was no statistically significant NR intervention effect found in this study.