Using a crossed laser-molecular beam scattering apparatus and tunable photoionization detection, these experiments determine the branching to the product channels accessible from the 2-hydroxyethyl radical, the first radical intermediate in the addition reaction of OH with ethene. Photodissociation of 2-bromoethanol at 193 nm forms 2-hydroxyethyl radicals with a range of vibrational energies, which was characterized in our first study of this system (J. Phys. 10 Chem. A 2010, 114, 4934). In this second study, we measure the relative signal intensities of ethene (at m/e = 28), vinyl (at m/e = 27), ethenol (at m/e = 44), formaldehyde (at m/e = 30), and acetaldehyde (at m/e = 44) products and correct for the photoionization cross sections and kinematic factors to determine a 0.765:0.145:0.026:0.063:<0.01 branching to the OH + C2H4, H2O + C2H3, CH2CHOH + H, H2CO + CH3, and CH3CHO + H product asymptotes. The detection of the H2O + vinyl product channel is surprising when starting from the CH2CH2OH radical adduct; prior studies had assumed that the H2O + vinyl products were solely from the direct abstraction channel in the bimolecular collision of OH and ethene. We suggest that these products may result from a frustrated dissociation of the CH2CH2OH radical to OH + ethene in which the C-O bond begins to stretch, but the leaving OH moiety abstracts an H atom to form H2O + vinyl. We compare our experimental branching ratio to that predicted from statistical microcanonical rate constants averaged over the vibrational energy distribution of our CH2CH2OH radicals. The comparison suggests that a statistical prediction using 1-D Eckart tunneling underestimates the rate constants for the branching to the product channels of OH + ethene, and that the mechanism for the branching to the H2O + vinyl channel is not adequately treated in such theories.