A study was undertaken to assess any differences between physicians' and nurses' estimates of burn size from drawings of 10 hypothetical patients with burns. Patient drawings were sent to the 199 burn facilities that are listed in the American Burn Association's Burn Care Resources in North America 1989-1990. The mailings were randomized between physicians and nurses. In order to have an even distribution, the cover letter asked the individual who received the forms to fill them out himself or herself rather than to pass them on to someone else. Forty-one nurses and 16 doctors (29%) returned the completed forms. The length of experience in burn care for the nurses averaged 9.3 years (range, 2 to 20 years), and the length of experience in burn care for the physicians averaged 14.7 years (range, 1 to 30 years [p = 0.01]). There were significant differences in the estimate of mean burn size between the physicians and nurses for only the two smallest drawings; in these cases, size may have been hard to translate from the patient to the drawing. There were four smaller burn charts (less than 20% body surface area), and in two (50%) of these, there were significant differences in the estimation of the burn size. For the six larger burn charts (greater than 20% body surface area), there were no differences in the size estimates. With regression analysis, no trend was noted that correlated the size estimates with the length of experience in burn care for either the physicians or the nurses. Additionally, the participants were asked to indicate which method of estimating burn size was used in their units. Thirty-five percent use the Rule of Nines, 33% use the Lund & Browder chart, 5.3% use Berkow's method, 3% use other methods, and 1.75% use a combination of the Lund & Browder chart and the Rule of Nines. Twenty-one percent failed to indicate any preferred method. Estimates of burn sizes with the use of standardized burn forms were consistently the same whether estimates were made by physicians or nurses. Sixty-eight percent of the respondents used the Rule of Nines or the Lund & Browder chart to estimate burn size, but 21% of the respondents failed to answer the question about which method is used in their units. In conclusion, there appears to be little variance in estimation of burn size as made by experienced burn nurses and physicians, and use of these estimates in a centralized data bank should be reliable.