Clinical Gait Analysis (CGA) is a pivotal technique for evaluating pathological conditions, particularly musculoskeletal disorders. However, its efficacy is often hindered by the fact that normative gait data is almost always used worldwide as a basis for CGA, regardless of differences in critical parameters such as BMI, age, gender, and walking speeds. To address this, we developed multiple regression models for predicting lower limb sagittal kinematic waveforms. We recorded anthropometric, demographic, spatiotemporal, and kinematic data from 30 healthy individuals. Leveraging the gait cycle time and joint angles as dependent variables, and BMI, age, gender, and walking speeds as predictors, we developed 46 regression equations. We employed PCHIP utilizing 80% of the kinematic data to reconstruct the waveforms and validated via leave-one-out cross validation. Our models successfully reconstructed hip, knee, and ankle kinematic waveforms, achieving R 2 ≥ 0.9 and RMSE ≤ 6° from the validation study. P-values < 0.05 as well as the clinical relevance of the predictors were considered during the regression analysis. These outcomes underscore the potential for our approach to be used as the basis to enhance the precision of region-specific gait data predictions, thus facilitating more accurate CGA.