Previous research showed that the most precise estimates of stink bug damage in developing cotton bolls are obtained by collecting soft quarter-sized bolls and dissecting them for signs of internal feeding damage, such as internal boll wall warts and/or stained lint. However, this method requires considerable time and effort; therefore, scouts and growers are unwilling to invest adequate resources to make sound pest management decisions. Here, the authors evaluated enumeration of external feeding lesions on groups of 10, 15, 20, or 25 bolls per sample as an alternative sampling procedure. Results relate the similarity of external boll feeding lesions to internal damage as a function of boll sample size and external lesion tally. Inverse prediction confidence intervals (CIs) were also calculated to predict internal boll damage on a new sample based only on external feeding lesions. Results show that linear regression model fit increased when examining at least 20 bolls per sample, and only one external lesion per boll provided as good of model fit as using a minimum of two, three, or four lesions per boll. Inverse prediction CIs suggested that more than one sample (20 bolls per sample) will be required to make external lesions an acceptable method for making reasonably accurate management decisions. F-test lack of fit and significant regression models suggest that examination of external lesions is a promising a method for estimating stink bug damage in cotton.