Automated reporting of estimated GFR (eGFR) with serum creatinine measurement is now common. We surveyed nephrologists in four countries to determine whether eGFR reporting influences nephrologists' recommendations for dialysis initiation. Respondents were randomly allocated to receive a survey of four clinical vignettes that included either serum creatinine concentration only or serum creatinine and the corresponding eGFR. For each scenario, the respondent was asked to rank his or her likelihood of recommending dialysis initiation on a modified 8-point Likert scale, ranging from 1 ("definitely not") to 8 ("definitely would"). Analysis of the 822 eligible responses received showed that the predicted likelihood of recommending dialysis increased by 0.55 points when eGFR was reported (95% confidence interval, 0.33 to 0.76), and this effect was larger for eGFRs >5 ml/min per 1.73 m(2) (P<0.001). Subgroup analyses suggested that physicians who had been in practice ≥13 years were more affected by eGFR reporting (P=0.03). These results indicate that eGFR reporting modestly increases the likelihood that dialysis is recommended, and physicians should be aware of this effect when assessing patients with severe CKD.