Objective To present for black and white patients with medically uncontrolled glaucoma 10-year results of treatment with 1 of 2 randomly assigned surgical intervention sequences. Design Randomized clinical trial. Participants Three hundred thirty-two black patients (451 eyes) and 249 white patients (325 eyes). Eyes had glaucoma that could not be controlled with medications alone. Methods Eyes were randomly assigned to 1 of 2 sequences: argon laser trabeculoplasty (ALT)-trabeculectomy-trabeculectomy (ATT) or trabeculectomy-ALT-trabeculectomy (TAT). Second and third interventions were offered after failure of the preceding intervention. Minimum required intraocular pressure (IOP) for intervention failure ranged upward from 18 mmHg, the value depending on whether recent optic disc or visual field (VF) deterioration occurred, and on the magnitude of the field defect. Patients were observed every 6 months, with total potential follow-up ranging from 8 years, 4 months to 13 years. Main outcome measures The averages over follow-up of (1) the percentage of eyes having moderate loss of VF and (2) the percentage of eyes having moderate loss of visual acuity (VA). Results Race-treatment interactions in VF and VA loss are significant for the 2 main outcome measures; therefore, results of treatment sequence differences are presented by race. In black patients the average percent of eyes with VF loss was less in the ATT sequence than in the TAT sequence, a difference that is not statistically significant at any visit. In white patients, conversely, after 18 months the average percent of eyes with VF loss was less in the TAT sequence, a difference that increases and is statistically significant in years 8 to 10. In both black and white patients, the average percent of eyes with VA loss was less in the ATT sequence; this difference is statistically significant throughout 10 follow-up years in black patients and is statistically significant only for the first year in white patients. In both black and white patients, average IOP reductions were greater in the TAT sequence, though the TAT-ATT difference was substantially greater in white patients. In both black and white patients, first-intervention failure rates were substantially lower for trabeculectomy than for trabeculoplasty. Ten-year cumulative incidence of unilateral VF impairment comparable to legal blindness was modest in eyes of black (ATT 11.9%, TAT 18.5%) and white (ATT 9.9%, TAT 7.3%) patients. Conclusions Although IOP was lowered in both sequences in black and white patients with medically uncontrolled glaucoma, long-term visual function outcomes were better for the ATT sequence in black patients and better for the TAT sequence in white patients.