Observational efforts in the last decade suggest the prevalence of photochemical hazes in exoplanetary atmospheres. Recent JWST observations raise growing evidence that exoplanetary hazes tend to have reflective compositions, unlike the conventionally assumed haze analogs, such as tholin and soot. In this study, I propose a novel hypothesis: diamond formation through chemical vapor deposition (CVD) may be happening in exoplanetary atmospheres. Using an aerosol microphysical model combined with the theory of CVD diamond and soot formation established in the industry community, I study how the haze composition evolves in exoplanetary atmospheres for various planetary equilibrium temperatures, atmospheric metallicity, and C/O ratio. I find that CVD diamond growth dominates over soot growth in a wide range of planetary parameters. Diamond haze formation is most efficient at T
eq ∼ 1000 K and low atmospheric metallicity ([M/H] ≤ 2.0), while soot could be the main haze component only if the atmosphere is hot (T
eq ≳ 1200 K) and carbon rich (C/O > 1). I also compute transmission, emission, and reflected light spectra, thereby suggesting possible observational signatures of diamond hazes, including the 3.53 μm feature of hydrogenated diamonds, anomalously faint thermal emission due to thermal scattering, and a drastic increase in geometric albedo. This study suggests that warm exoplanetary atmospheres may be favorable sites for forming CVD diamonds, which would be testable by future observations by JWST and Ariel as well as haze synthesis experiments under hot hydrogen-rich conditions.