The aim of this study was to conduct an economic evaluation for the treatment of subjective tinnitus using different modalities of cognitive behavioral therapy (CBT) in Australia. A decision tree model was used to conduct a cost-utility analysis for CBT to determine the cost effectiveness for tinnitus treatments, in terms of cost per responder and cost per quality-adjusted life-year (QALY), from a health system perspective using a 2-year time horizon. Meta-analysis was used to differentiate the levels of effectiveness between three delivery methods for CBT: individual face-to-face care (fCBT), group sessions (gCBT), and a supported internet program (iCBT). One-way sensitivity analysis and probabilistic sensitivity analysis (PSA) explored the uncertainty surrounding model inputs and outcomes. Results were presented as incremental cost-effectiveness ratios compared with no treatment, and as net monetary benefit at a $50,000 willingness-to-pay threshold. Compared with no treatment, the incremental cost per responder was $700 for gCBT, $871 for iCBT, and $1380 for fCBT. The base case incremental cost-effectiveness ratio was $35,363 per QALY for fCBT, $17,935 per QALY for gCBT, and $22,321 per QALY for iCBT compared with no treatment, although there was substantial uncertainty around the QALY gain for responders. Net monetary benefit was $356 (fCBT), $555 (gCBT), and $487 (iCBT), indicating the treatments were cost effective compared with no treatment. One-way sensitivity analysis revealed the results were most sensitive to the probability of a positive response to treatment and treatment length. The PSA found the probability of being cost effective compared with no treatment for gCBT was 99.8%, iCBT 98.4%, and fCBT 71.5% at a willingness-to-pay of $50,000 per QALY, although QALY gain remained at a fixed value in the PSA. CBT for tinnitus was likely to be cost effective compared with no treatment regardless of treatment modality, assuming they are not mutually exclusive. Of the interventions, gCBT was the lowest cost per responder and lowest cost per QALY. Internet CBT obtained comparable economic outcomes due to similar treatment effectiveness and cost. Group CBT and iCBT warrant greater adoption in clinical practice for the treatment of subjective tinnitus. Further research on preference-based utility measures for varying levels of tinnitus severity and the durability of treatment effect is required to enhance the quality of economic evaluation in this field.