The doubly labelled water method is a common means of investigating field metabolic rates (FMRs) of free-ranging animals by injecting oxygen and hydrogen isotopes. Compared with a general two-sample approach including double blood sampling, a single-sample approach, which includes an estimation of initial isotope enrichment and single blood sampling, has been developed as a less invasive technique with lower impact on the behavior of study subjects. However, little attempt has been made to improve the indirect estimation of initial isotope enrichment and to apply the two-pool model for calculating FMR from the single-sample approach. Therefore, we studied the validity of a single-sample approach in the Streaked Shearwater Calonectris leucomelas. We developed equations for estimating initial isotope enrichment based on the amount of injected isotopes and body mass collected from 15 shearwaters. Then, for six shearwaters subjected to a two-sample approach, we calculated the turnover rates of oxygen and hydrogen isotopes (ko and kd), and FMR using the two-pool model with measured and initial isotope enrichments. The arithmetic errors were -0.01% for the estimated initial enrichments of oxygen isotope and -0.11% for hydrogen isotope. The ko, using estimated initial isotope, is overestimated by 3.2% on average, while kd is underestimated by 0.4% in comparison with those measured by the two-sample approach. The FMR measured by the single-sample approach are overestimated by 12.0% (± 12.1 SD) in comparison with those measured by the two-sample approach. We were able to estimate reliably the initial enrichments of both isotopes and apply the two-pool model in the calculation of FMR.