The efficacy of a research-based number-sense intervention for low-achieving kindergartners was examined. Children (N = 126) were randomly assigned to 1 of 3 conditions: a number-sense intervention followed by a number-fact practice session, an identical number-sense intervention followed by a number-list practice session, or a business-as-usual control group. The interventions were delivered in a small-group setting over 24 half-hour lessons. Both intervention groups performed better than controls on measures of number sense, arithmetic fluency, and general mathematics calculation achievement at immediate posttest. However, the number-fact practice condition gave children an additional advantage over the number-list practice condition on the outcomes at delayed posttest 8 weeks later. The number-fact practice condition was especially effective for producing gains in English learners.