We estimate a model of household use of alternative payment instruments (cash and bank cards) using a new dataset from a survey of Russian households. In our model, household preferences are determined by the perceived attributes of these instruments, and hence their choice regarding payment methods depends on the differences in these characteristics across these instruments. The results indicate that consumer choice is sensitive to the perceived characteristics at statistically significant level. We employ the estimated model to evaluate demand for central bank digital currency (CBDC), based on its expected design and consumer perceptions of it. We discuss several illustrative projections to demonstrate the application of the tool developed. The predicted use of CBDC varies considerably, depending on the perceived characteristics, although the projected use of CBDC in household transactions is limited under conservative assumptions: the projected share of digital rubles in the total number of transactions varies from 18 percent to 35 percent.