Estimation of soil texture is an essential skill for soil scientists, but a difficult skill to learn. We compared two approaches for teaching soil texture: the traditional flowchart approach and an experimental self-calibration method. Three lab sections of an introductory soil science class were assigned to learn by the flowchart method (n = 60) and three lab sections were assigned to learn by the self-calibration method (n = 53). Students who learned by self-calibration recorded data for observations related to clay (i.e., soil cohesion and plasticity) and sand (i.e., volume remaining after separating sand by washing) and instructors used this data to provide each student with a personalized method. Students who learned by the flowchart method practiced following a protocol by molding the soil into a ribbon to estimate clay and estimating sand by feel. After spending one lab period learning their assigned methods, all students attempted to identify the texture of four soils. The accuracy of texture class assignments was slightly higher for the self-calibration group (27%) compared to the flowchart group (20%). Students in the self-calibration group performed better at estimating clay because they were directed to use the method that worked best for them. However, the flowchart students performed better at sand estimation because feel is an easier technique to learn compared to separating sands by washing. The self-calibration approach is recommended for advanced soil science courses focused on field methods, with modification to incorporate feel as a metric for sand estimation.