Carpal tunnel syndrome (CTS) is highly prevalent, resulting in decreased function and increased need for costly healthcare services. External wrist ratio (depth/width >0.70) is a strong predictor of the development of CTS and has been suggested to be a proxy for internal carpal tunnel (CT) shape. Conversely, sonography can more directly evaluate CT shape. The purpose of our study was to explore the relationship between wrist ratio and sonographic CT measurements to (1) evaluate the reliability of sonographic CT measurements and (2) explore how external wrist measures relate to anthropometric features of the CT. We used sonographic imaging on a sample of healthy participants (n = 226) to measure CT cross‐sectional area, depth, width, and depth/width ratio. We conducted exploratory correlation and regression analyses to identify relationships of these measures with external wrist ratio. Reliability for dominant and nondominant sonographic CT measures ranged from good to excellent (0.79–0.95). Despite a moderate correlation between CT width and depth and their external wrist counterparts (0.33–0.41, p < 0.001), wrist ratio and CT ratio demonstrated weak to no correlation (dominant: r = 0.12, p = 0.053; nondominant: r = 0.20, p = 0.002) and the mean CT ratio was far lower than the mean wrist ratio (0.45 vs. 0.71 bilaterally). Supporting this, we observed several key differences in the relationship between external wrist measures compared to corresponding CT measures. Additionally, regression analyses combining participant factors and CT measurements produced models accounting for less than 15% of the variability in external wrist ratio (linear models) or correctly predicting less than 68% of wrist ratio‐based risk categorization (logistic models). Overall, among healthy young adults, wrist shape is not an adequate proxy for CT shape.