The determinants of referral for transplantation in women have not been well-studied. Similarly, factors determining survival on dialysis and the rate of transplantation in women remain controversial. Women have been reported to have lower rates of transplantation than men, and black women have the lowest rates of all groups. We questioned whether black women were referred at lower rates than whites and if race and other socioeconomic factors predicted referral, rate of renal transplantation, and patient survival on dialysis. All women in Allegheny or Philadelphia counties in Pennsylvania initiating dialysis between January 1, 1990, and December 31, 1992, were eligible for this study. Information was requested by questionnaire from each dialysis unit in these areas. Of the 383 eligible patients, completed questionnaires were obtained for 276 (72%). Ninety-three (54.7%) of the black patients and 57 (53.8%) of the white patients were referred for transplantation (P = 0.8). Declining the transplant option was the most common reason for nonreferral in both races. Patients with high school or greater education were approximately twice as likely to be referred than those with grade school educations (odds ratio [OR], 2.2; P = 0.04). Patients with coexisting illness were 67% (OR, 0.33; P = 0.004) less likely to be referred compared with patients with no other illnesses. Each additional year of age reduced the chances of being referred by 6% (OR, 0.94; P = 0.0001). Homemakers were 83% (P = 0.0008) and others 55% (P = 0.07) less likely to be referred compared with employed patients. Patients in Philadelphia County were 56% less likely to be referred compared with those in Allegheny County (P = 0.024). Race was not significantly associated with referral. Predictors of transplantation included age (RR, 0.96; confidence interval [CI], 0.93 to 0.99; P = 0.13), white race (RR, 2.2; CI, 1.3 to 4.0; P = 0.0053), presence of other illnesses (RR, 0.37; CI, 0.21 to 0.65; P = 0.0006), and employment status. White homemakers were 86% (RR, 0.14; CI, 0.03 to 0.6; P = 0.0082) less likely than those with other employment situations to receive a transplant. Factors predicting patient survival on dialysis included race, educational status, and presence of comorbid illnesses. White patients were approximately four times (RR, 3.7; CI, 1.7 to 8.1; P = 0.002) more likely to die than black patients. Patients with high school or greater education were 56% (RR, 0.44; CI, 0.2 to 0.92; P = 0.008) less likely to die than those with grade school education alone. Patients with at least one coexisting illness were approximately 1.7 times (RR, 1.68; CI, 1.1 to 2.4; P = 0.001) more likely to die than those without other illnesses. In summary, race was not a factor in referral for transplantation, but was predictive of transplantation and patient survival on dialysis. Socioeconomic factors such as educational status, age, and employment status were highly predictive of transplantation and long-term survival on hemodialysis. White homemakers unexpectedly received transplants less than any other group of dialysis patients. Further study is needed to determine why these potential transplant patients have declined or deferred transplantation.