AbstractIn matching experimental data to kinetic models for hydrogenation of methylcis‐9,cis‐15‐linoleate, ethylene glycol diundecylenate and 1,7‐octadiene with platinum or palladium catalysts, only if a diene‐to‐saturate shunt of 15‐ to 25% is included can a good fit be secured between the model and experimental composition. With nickel the shunt is much smaller or absent. Although the shunt might be attributed most readily to a liquid diffusion effect, its independence of temperature, viscosity and presence or absence of solvents and catalyst support suggests that some other factor is responsible. Evidence for similar shunts is found for other materials, including soybean triglycerides.