Abstract We report the early discovery and multicolor (BVI) high-cadence light-curve analyses of the rapidly declining sub-Chandrasekhar Type Ia supernova KSP-OT-201509b (= AT 2015cx) from the KMTNet Supernova Program. The Phillips and color stretch parameters of KSP-OT-201509b are ΔM _B,15 ≃ 1.62 mag and s _BV ≃ 0.54, respectively, at an inferred redshift of 0.072. These, together with other measured parameters (such as the strength of the secondary I-band peak, colors, and luminosity), identify the source to be a rapidly declining Type Ia of a transitional nature that is closer to Branch-normal than 91bg-like. Its early light-curve evolution and bolometric luminosity are consistent with those of homologously expanding ejecta powered by radioactive decay and a Type Ia SN explosion with 0.32 ± 0.01 M _⊙ of synthesized ⁵⁶Ni mass, 0.84 ± 0.12 M _⊙ of ejecta mass, and (0.61 ± 0.14) × 10⁵¹ erg of ejecta kinetic energy. While its B − V and V − I colors evolve largely synchronously with the changes in the I-band light curve, as found in other supernovae, we also find the presence of an early redward evolution in V − I prior to −10 days since peak. The bolometric light curve of the source is compatible with a stratified ⁵⁶Ni distribution extended to shallow layers of the exploding progenitor. Comparisons between the observed light curves and those predicted from ejecta–companion interactions clearly disfavor Roche lobe–filling companion stars at large separation distances, thus supporting a double-degenerate scenario for its origin. The lack of any apparent host galaxy in our deep stack images reaching a sensitivity limit of ∼28 mag arcsec⁻² makes KSP-OT-201509b a hostless Type Ia supernova and offers new insights into supernova host galaxy environments.