As a major global health concern, coronary artery disease (CAD) demands precise, noninvasive diagnostic methods like cardiopulmonary exercise testing (CPET) for effective assessment and management, balancing the need for accurate disease severity evaluation with improved treatment decision-making. Our objective was to develop and validate a nomogram based on CPET parameters for noninvasively predicting the severity of CAD, thereby assisting clinicians in more effectively assessing patient conditions. This study analyzed 525 patients divided into training (367) and validation (183) cohorts, identifying key CAD severity indicators using least absolute shrinkage and selection operator (LASSO) regression. A predictive nomogram was developed, evaluated by average consistency index (C-index), the area under the receiver operating characteristic curve (AUC), calibration curve, and decision curve analysis (DCA), confirming its reliability and clinical applicability. In our study, out of 25 variables, 6 were identified as significant predictors for CAD severity. These included age (OR = 1.053, P < .001), high-density lipoprotein (HDL, OR = 0.440, P = .002), hypertension (OR = 2.050, P = .007), diabetes mellitus (OR = 3.435, P < .001), anaerobic threshold (AT, OR = 0.837, P < .001), and peak kilogram body weight oxygen uptake (VO2/kg, OR = 0.872, P < .001). The nomogram, based on these predictors, demonstrated strong diagnostic accuracy for assessing CAD severity, with AUC values of 0.939 in the training cohort and 0.840 in the validation cohort, and also exhibited significant clinical utility. The nomogram, which is based on CPET parameters, was useful for predicting the severity of CAD and assisted in risk stratification by offering a personalized, noninvasive diagnostic approach for clinicians.