Declining trust in statistical agencies has recently complicated the endeavour to collect high-quality, timely data that are used to inform US policy and practice. Given this context, understanding how respondents choose to trust particular statistical agencies and their products is incredibly important. This article details a series of cognitive interviews (85) and focus groups (3) used to measure how the US public develops trust for statistical agencies, their statistical products and their use of administrative records. Results show that respondents use two models of trust in their rationale: experience based and cultural‒repertoire based. When respondents did not have experience with a particular institution and/or its product, cultural values including personal liberty, cost-savings and the promotion of social goods (for example, government-sponsored schools and hospitals) were found to influence their motivations to trust or distrust. As a result, appeals to cultural values may have the potential to increase trust among respondents. Familiarity with statistical agencies and their products may also increase respondents’ levels of trust.