The paper investigates the relationship between sectoral capital–labor ratios and total factor productivity (TFP) in the context of the Balassa–Samuelson model. It is shown that, under certain assumptions, the model implies that both traded‐ and nontraded‐goods sectors’ capital–labor ratios should be cointegrated with the traded‐goods sector’s TFP. Evidence from an intersectoral database for 14 OECD countries broadly supports this implication of the model. In addition to shedding light on the evolution of sectoral capital–labor ratios, the results also alleviate concerns regarding the reliability of capital stock data.