Culture has long shaped individuals’ response to problems. A classic puzzle in the sociology of health and illness is discrepancy between theory and research regarding cultural beliefs and medical care service use. “Utilization research,” examining individuals’ responses to the onset of health problems, has not consistently affected culture on the uptake of formal treatment. While ethnographic research often describes how culture shapes illness behaviors, survey-based studies rarely find significant effects of predispositions once “need” is controlled. In quantitative studies, individuals report supportive treatment beliefs or predispositions to use services but low utilization levels, reinforcing claims about lack of utility of cultural ideologies in health-care decision making. We ask whether innovations in the sociology of culture and cognition provide the theoretical scaffolding to conceptualize and measure culture in health service utilization. Examining data from the General Social Survey, we focus on how approaches to culture might explain the paradox of high cultural predispositions and low actual use. Children with mental health problems provide a comparison between suggestions and endorsements. Suggestions, sources of care offered by individuals in response to a case description without any other social cues, align with new cultural approaches, and are measured by responses to open-ended questions about what should be done for the child described meeting clinical criteria f or ADHD, major depression, asthma, or “daily troubles”). Endorsements, requiring less cognitive work and cultural resistance, align with traditional conceptualizations of culture, and are measured by closed-ended questions that ask respondents to agree or disagree with seeking help from different treatment options placed later in the survey. We find that suggestions reveal cultural predispositions to use services corresponding closely to reported utilization levels; endorsements reveal high, unrealistic cultural predispositions to use services. Further, suggestions are associated with sociodemographics that proxy culture (e.g., race), while endorsements are associated only with perceived need.