Although Indonesia possesses 20% of the world’s mangroves (3.36 million hectares), these ecosystems face rapid degradation due to land conversion across Sumatra, Kalimantan, Java, and Sulawesi. Mangrove rehabilitation efforts have been slow, with only 130,000 hectares rehabilitated by 2024 against a 600,000-hectare target. Moreover, standardized multi-site carbon stock data from rehabilitated mangroves remain limited due to most research efforts involving single-location studies. This study evaluates carbon stock potential across five rehabilitation sites under World Wildlife Fund Indonesia programs: Pati, Situbondo, Jember, Tarakan, and Jeneponto. Field surveys were conducted across these five distinct sites utilizing a cluster sampling design. A total of 14 plots were established, comprising 10 plots within the rehabilitation zones and 4 external comparison plots. To capture the primary data, these plots were further subdivided into 47 subplots, of which 34 were internal and 13 external. Vegetation and necromass were inventoried using circular plot methods. For soil analysis, sampling was performed using a soil auger at vertical depth intervals ranging from 5 to 200 cm. Soil organic matter was subsequently determined via the loss on ignition method, and total soil carbon stocks were calculated by integrating bulk density, carbon fraction, and the respective horizon depths. The Mann‑Whitney pairwise comparisons were applied for total carbon stock comparison among sites. Correction for multiple comparisons was performed using the Bonferroni correction method (28 pairwise tests, adjusted α = 0.05/28≈0.00179). The results indicated substantial variation in carbon stocks among rehabilitation sites, ranging from 258.64±105.78 to 802.42±72.00 MgC ha-1. The highest carbon stock was recorded in Pati, which was dominated by sediment carbon at 764.67±69.04 MgC ha-1, while the lowest was observed in Jember. Sediment carbon represents the primary carbon storage component (>80% of total stock). Differences in carbon stocks across locations were not associated with ecosystem conditions or the age of rehabilitation; in contrast, sites with longer rehabilitation periods generally exhibited higher vegetation biomass. Across all rehabilitation sites, necromass contribution remained low (<5%), with its highest density of 4.55±5.87 MgC ha-1 recorded in Tarakan. These findings provide a scientific basis for further monitoring and the improvement of mangrove rehabilitation programs by responsible stakeholders.