Coronary slow flow (CSF) is a microvascular disease characterized by delayed opacification of the epicardial coronary arteries during angiography. The main pathogenesis of CSF is endothelial dysfunction caused by diffuse atherosclerosis. Dyslipidemia is one of the primary factors raising the risk of atherosclerosis. Compared to conventional lipid profiles, non-traditional lipid profiles more accurately reflect dyslipidemic status. In this work, we compared the non-high density lipoprotein-cholesterol (HDL-C)/HDL-C ratio (NHHR) with other conventional and non-conventional lipid profiles in order to determine its impact on CSF. A total of 9112 subjects who underwent coronary angiography were screened retrospectively, of whom 130 subjects with CSF and 130 subjects with normal CF were included. Multivariate regression analysis was used to identify independent predictors of CSF. Additionally, in order to predict CSF, the diagnostic accuracies of NHHR and other non-traditional lipid profiles were examined. There were significantly higher non-traditional lipid profiles in the CSF group (all p < 0.001). Compared to other non-traditional lipid profiles, NHHR had a stronger association with thrombolysis in myocardial infarction frame count (r = 0.3593, p < 0.0001). In addition to NHHR, non-HDL-C, Castelli's risk index-II, atherogenic index of plasma, plasma glucose, dyslipidemia, smoking, and body mass index were identified as independent predictors of CSF. The ability of NHHR to detect CSF was superior to other non-traditional lipid profiles (area under the curve: 0.785; confidence interval: 0.730-0.840; p < 0.001). NHHR was found to be a potent and reliable predictor of CSF. This indicates that NHHR can be used as a reliable biomarker for risk stratification of CSF.