Tree and log diameters are usually measured outside bark, but inside-bark diameters are of greater economic interest and are often derived with local or regional bark thickness equations. To date, the influence of measurement method, sampling design, and sample size on bark thickness equation accuracy and precision has received limited attention. The objectives of this study were to use an extensive regional bark thickness dataset for Norway spruce (Picea abies (L.) Karst) in southwestern Germany to (1) quantify the accuracy and precision of bark thickness measurements with a Swedish bark gauge, (2) determine the required number of measurements to assess the within-tree variation, and (3) estimate the required sample sizes per plot and per region to develop an accurate bark thickness prediction equation. Bark gauge readings were validated with measurements derived from X-ray computed tomography (CT) and indicate that Swedish bark gauges generally overestimated bark thickness by 13.6% ± 28.4% (mean ± standard deviation). Results suggested having at least one measurement location every 2 m along a tree bole and at least five bark thickness measurements per each of these locations to achieve an allowable error of <15%. For the study area, Monte Carlo simulations indicated that a total sample size of 50–250 trees was needed, depending on the complexity of the desired bark thickness model. Overall, this analysis indicated that there was relatively high within- and between-tree variation in bark thickness, but adequate sampling methods and sample sizes produced highly accurate bark thickness equations.