Abstract The three-rung distance ladder, which calibrates Type Ia supernovae (SNe Ia) through stellar distances linked to geometric measurements, provides the highest precision direct measurement of the Hubble constant. In light of the Hubble tension, it is important to test the individual components of the distance ladder. For this purpose, we report a measurement of the Hubble constant from 35 extragalactic Cepheid hosts measured by the SH0ES team, using their distances and redshifts at cz ≤ 3300 km s⁻¹, instead of any more distant SNe Ia, to measure the Hubble flow. The Cepheid distances are calibrated geometrically in the Milky Way, NGC 4258, and the Large Magellanic Cloud. Peculiar velocities are a significant source of systematic uncertainty at z ∼ 0.01, and we present a formalism for both mitigating and quantifying their effects, making use of external reconstructions of the density and velocity fields in the nearby universe. We identify a significant source of uncertainty originating from different assumptions about the selection criteria of this sample, whether distance or redshift limited, as it was assembled over three decades. Modeling these assumptions yields central values ranging from H ₀ = 71.7 to 76.4 km s⁻¹ Mpc⁻¹. Combining the four best-fitting selection models yields H 0 = 72.9 − 2.2 + 2.4 as a fiducial result, at 2.4σ tension with Planck. While SNe Ia are essential for a precise measurement of H ₀, unknown systematics in these supernovae are unlikely to be the source of the Hubble tension.