Military service members (SMs) with mild traumatic brain injury (mTBI or concussion) frequently report cognitive and behavioral difficulties. Currently, military clinical guidelines recommend clinician-run, manualized cognitive rehabilitation (CR) to treat these symptoms; however, it is unclear whether this approach adequately addresses the unique needs of warfighters. Computerized cognitive training (CCT) programs represent an innovative, promising approach to treating cognitive difficulties; however, whether these programs can effectively remediate cognitive impairment in individuals with mTBI remains unclear. A total of 65 SMs with a history of at least 1 diagnosed mTBI were recruited from a military hospital. Participants received 1 of 2 interventions: Clinician-run, manualized CR (Study of Cognitive Rehabilitation Effectiveness [SCORE]; n = 37), consisting of 60 total intervention hours over 6 weeks, or CCT (n = 28), in which participants trained with either a commercial CCT (n = 14) or noncommercial CCT (n = 14), for a total of 12 hours over 4 weeks. Participants were assessed pre- and postintervention, using a combination of self-report and objective outcome measures: Key Behaviors Change Inventory (KBCI), a self-report measure of functional difficulties; Paced Auditory Serial Addition Test (PASAT), an objective cognitive assessment that measures both information processing speed and sustained and divided attention; and Symbol Digit Modalities Test (SDMT), an objective cognitive assessment that measures information processing speed. Mixed ANOVA revealed no interaction effect between intervention type and time (pre- and postassessment) on the PASAT (P = .643, ηp2 = 0.003), SDMT (P = .423, ηp2 = 0.010), or KBCI (P = .434, ηp2 = 0.010); however, there was a significant within-group main effect (time) on all 3 outcome measures (PASAT P < .001, ηp2 = 0.54; SDMT P < .001, ηp2 = 0.25; and KBCI P = .001, ηp2 = 0.15). On average, participants showed improvement over baseline on the PASAT (SCORE delta = 6.98, SD = 7.25, P < .001; CCT delta = 7.79, SD = 6.45, P < .001), SDMT (SCORE delta = 4.62, SD = 8.82, P = .003; CCT delta = 6.58, SD = 10.81, P = .003), and KBCI (SCORE delta = -3.22, SD = 7.09, P = .009; CCT delta = -2.00, SD = 4.72, P = .033). Additional analysis comparing the relative effectiveness of the 2 different CCTs revealed that while training with either program resulted in improved performance on the PASAT (P < .001, ηp2 = 0.627), SDMT (P = .003, ηp2 = 0.286), and KBCI (P = .036, ηp2 = 0.158), there was no interaction effect of CCT program type and change over time for any measure (PASAT P = .102, ηp2 = 1.00; SDMT P = .317, ηp2 = 0.038; and KBCI P = .719, ηp2 = 0.005). We showed that CCT programs do not differ in efficacy compared to clinician-run, manualized CR for treating symptoms associated with mTBI; however, exploratory analyses suggest that each approach may have distinct advantages for treating specific symptoms. Additionally, we showed that the improvement in the CCT intervention did not differ between those who trained using the commercial program vs. those who trained with the noncommercial program.