Cardiovascular disease (CVD) is a global health issue. Low- and middle-income countries (LMICs) are facing early CVD-related morbidity. Early diagnosis and treatment are an effective strategy to tackle CVD. The aim of this study was to assess the ability of community health workers (CHWs) to screen and identify persons with high risks of CVD in the communities, using a body mass index (BMI)-based CVD risk assessment tool, and to refer them to the health facility for care and follow-up. This was an action research study conducted in rural and urban communities, conveniently sampled in Rwanda. Five villages were randomly selected from each community, and one CHW per each selected village was identified and trained to conduct CVD risk screening using a BMI-based CVD risk screening tool. Each CHW was assigned to screen 100 fellow community members (CMs) for CVD risk and to refer those with CVD risk scores ≥10 (either moderate or high CVD risk) to a health facility for care and further management. Descriptive statistics with Pearson’s chi-square test were used to assess any differences between rural and urban study participants vis-à-vis the key studied variables. Spearman’s rank coefficient and Cohen’s Kappa coefficient were mainly used to compare the CVD risk scoring from the CHWs with the CVD risk scoring from the nurses. Community members aged 35 to 74 years were included in the study. The participation rates were 99.6% and 99.4% in rural and urban communities, respectively, with female predominance (57.8% vs. 55.3% for rural and urban, p-value: 0.426). Of the participants screened, 7.4% had a high CVD risk (≥20%), with predominance in the rural community compared to the urban community (8.0% vs. 6.8%, p-value: 0.111). Furthermore, the prevalence of moderate or high CVD risk (≥10%) was higher in the rural community than in the urban community (26.7% vs. 21.1%, p-value: 0.111). There was a strong positive correlation between CHW-based CVD risk scoring and nurse-based CVD risk scoring in both rural and urban communities, 0.6215 (p-value < 0.001) vs. 0.7308 (p-value = 0.005). In regard to CVD risk characterization, the observed agreement to both the CHW-generated 10-year CVD risk assessment and the nurse-generated 10-year CVD risk assessment was characterized as “fair” in both rural and urban areas at 41.6% with the kappa statistic of 0.3275 (p-value < 001) and 43.2% with kappa statistic of 0.3229 (p-value =0.057), respectively. In Rwanda, CHWs can screen their fellow CMs for CVD risk and link those with high CVD risk to the healthcare facility for care and follow-up. CHWs could contribute to the prevention of CVDs through early diagnosis and early treatment at the bottom of the health system.