I compared the effects on plasma corticosterone levels of two methods of collecting blood samples during standardized capture and handling stress protocols. In one method, individual nestling American Kestrels (Falco sparverius) were bled at three time periods: when initially removed from the nest, and 15 and 30 min later. In the other method, siblings removed from a nest were bled once each, either at the time of removal, or 15 or 30 min later. I found that there was no difference between the two groups in plasma corticosterone levels at the first sampling period, but 15 and 30 min after capture the singly-bled birds had significantly higher plasma corticosterone levels than the multiply-bled nestlings. The results suggest that data from multiplybled birds underestimate actual circulating hormone levels. The underlying mechanism for this phenomenon is unknown, although it may involve hemodilution.Key words: American Kestrel, stress, corticosterone, hemodilution, nestlings Organisms experiencing stressors typically secrete glucocorticoids to maintain or restore homeostasis (Wingfield and Romero 2001). The pattern of corticosterone secretion in response to a standardized stressor frequently is used to compare how different organisms respond to disruptive events. For birds, the most common technique is to use capture and handling stress to activate the hypothalamic-pituitary-adrenal (HPA) axis, with the sensitivity of the axis to the stressor determined through repeated blood sampling followed by radioimmunoassay (RIA) of the blood samples for corticosterone. Birds are held in a cloth bag between sampling sessions. This protocol facilitates comparison of adrenal responsiveness within and among species, and the results of such studies have enhanced our understanding of how ecological factors modulate stress responses (Wingfield et al. 1998, Wingfield andRomero 2001).The importance of a standardized approach in collecting blood samples to measure adrenoresponsiveness has been acknowledged, either explicitly or implicitly, in myriad papers. However, subtle, but possibly important, procedural differences occur among studies that use this protocol. For example, in many studies morphological measurements are taken after the last sampling period, so that the only time birds are handled is when blood samples are taken (Wingfield et al. 1992, Silverin et al. 1997, Blas et al. 2005. In contrast, sometimes morphological measurements are taken in between blood samples, when the birds would otherwise be in a cloth bag (Meddle et al. 2003, Adams et al. 2005. Similarly, occasionally birds are held in small cages in between sampling periods (e.g., Love et al. 2005) rather than in cloth bags. Furthermore, different trapping methods may be utilized in the same investigation (e.g., Lynn et al. 2003). In addition, in most studies similar amounts of blood are removed at each sampling period, but in others the initial blood sample may be larger, if additional hematological measurements are performed (e.g., Clinchy et al. 2004)....