Several studies have attempted to calculate the energy intake rates of shorebirds. One method of doing this involves estimating the size of prey items in relation to the bird's bill length, and then using an average bill length to convert this value to actual prey length in order to estimate prey mass/energy content. Although this technique is used regularly, no study has assessed under what conditions the method is reliable. Observer distance from the bird, prey type, size and orientation in the bill, bill length and shape, and light conditions may all affect the accuracy of the estimates. This study sought to establish the conditions under which the method is or is not reliable, using three different prey types (“thin” and “fat” polychaetes, and crabs) and a long- and short-billed bird. It was found that if observers were calibrated, and their results were corrected using regression analysis, errors in estimates of food intake were much reduced. The time allowed to view the prey significantly affected the error in the result, with shorter viewing time causing greater errors, and observations made into the light were less accurate than those made with the light behind the observer. Large prey types, or even larger size classes of the same species, are estimated less accurately than small prey, and accuracy decreases as prey size increases as multiples of bill length. If separate calibration regressions are generated for large and small prey, food intake estimates are more accurate than if a single regression is used incorporating the full range of prey sizes. Sexual dimorphism in bill length can also increase error in the results.