OBJECTIVES: Investigate the association between socioeconomic deprivation and completeness of cardiovascular disease (CVD) risk factor recording in primary care, uptake of screening in people with incomplete risk factor recording and with actual CVD risk within the screened subgroup. DESIGN: Cross-sectional study. SETTING: Nine UK general practices. PARTICIPANTS: 7987 people aged 50-74 years with no CVD diagnosis. METHODS: CVD risk was estimated using the Framingham equation from data extracted from primary care electronic health records. Where there was insufficient information to calculate risk, patients were invited to attend a screening assessment. ANALYSIS: Proportion of patients for whom clinical data were sufficiently complete to enable CVD risk to be calculated; proportion of patients invited to screening who attended; proportion of patients who attended screening whose 10-year risk of a cardiovascular event was high (>20%). For each outcome, a set of logistic regression models were run. Crude and adjusted ORs were estimated for person-level deprivation, age, gender and smoking status. We included practice-level deprivation as a continuous variable and practice as a random effect to account for clustering. RESULTS: People who had lower Indices of Multiple Deprivation (IMD) scores (less deprived) had significantly worse routine CVD risk factor recording (adjusted OR 0.97 (0.95 to 1.00) per IMD decile; p=0.042). Screening attendance was poorer in those with more deprivation (adjusted OR 0.89 (0.86 to 0.91) per IMD decile; p<0.001). Among those who attended screening, the most deprived were more likely to have CVD risk >20% (OR 1.09 (1.03 to 1.15) per IMD decile; p=0.004). CONCLUSIONS: Our data suggest that those who had the most to gain from screening were least likely to attend, potentially exacerbating existing health inequalities. Future research should focus on tailoring the delivery of CVD screening to ensure engagement of socioeconomically deprived groups.