Density functional theory (DFT) calculations and temperature programmed desorption (TPD) experiments were performed to study the adsorption of hydrogen on the Co(111) and Co(100) surfaces. On the Co(111) surface, hydrogen adsorption is coverage dependent and the calculated adsorption energies are very similar to those on the Co(0001) surface. The experimental adsorption saturation coverage on the Co(111)/(0001) surface is ¿max ˜ 0.5 ML, although DFT predicts ¿max ˜ 1.0 ML. DFT calculations indicate that preadsorbed hydrogen will kinetically impede the adsorption process as the coverage approaches ¿ = 0.5 ML, giving rise to this difference. Adsorption on Co(100) is coverage independent up to ¿ = 1.00 ML, contrasting observations on the Ni(100) surface. Hydrogen atoms have low barriers of diffusion on both the Co(111) and Co(100) surfaces. A microkinetic analysis of desorption, simulating the expected TPD experiments, indicated that on the Co(111) surface two TPD peaks are expected, while on the Co(100) only one peak is expected. Low coverage adsorption energies of between 0.97 and 1.1 eV are obtained from the TPD experiment on a smooth single crystal of Co(0001), in line with the DFT results. Defects play a important role in the adsorption process. Further calculations on the Co(211) and Co(221) surfaces have been performed to model the effects of step and defect sites, indicating that steps and defects will expose a broad range of adsorption sites with varying (mostly less favorable) adsorption energies. The effect of defects has been studied by TPD by sputtering of the Co crystal surface. Defects accelerate the adsorption of hydrogen by providing alternative, almost barrierless pathways, making it possible to increase the coverage on the Co(111)/(0001) surface to above ¿ = 0.50 ML. The presence of defects at a high concentration will give rise to adsorption sites with much lower desorption activation energies, resulting in broad low temperature TPD features.