Remote instruction is becoming increasingly common, yet few studies have directly compared remote and in-person instruction in a controlled manner. We used a reversal design to compare the effects of in-person and remote instruction for six preschool participants with disabilities learning tacts and sight words. Distribution of instruction, methodology, and materials across in-person and remote conditions were equated so that the only difference across conditions was the modality of instruction. Across conditions, we measured (1) the rate of learning; (2) the rate of trial presentation; (3) number of targets mastered; and (4) percentage of correct responses during follow-up assessment. Results indicate that three of six participants reliably met acquisition criteria and completed instruction faster in-person, with mixed results for the other three participants. No consistent difference was observed in response maintenance or generalization across modalities. These findings add to existing literature suggesting that remote instruction should be considered in situations where in-person instruction is unavailable.