Most left ventricular (LV) Doppler measurements vary significantly with age and gender, making it necessary to correct them for physiological variances. We aimed to verify the hypothesis that different Doppler measurements correlate nonlinearly with different biometric variables raised to different scaling factors and exponents. In this work, a total of 23 LV Doppler parameters were measured in 1224 healthy Chinese adults. An optimized multivariable allometric model (OMAM) and scaling equations were developed in 70% of the subjects (group A), and the reliability of the model and equations was verified using the remaining 30% of the subjects (group B) as well as 183 overweight subjects (group C). The single-variable isometric model (SVIM) with body surface area (BSA) as a scaling variable was used for comparison. Before correction, all 23 LV Doppler parameters correlated significantly with one or more of the biometric variables. In group B, gender differences were found in 47.8% (11/23) of the parameters and were eliminated in 81.8% (9/11) of the parameters after correction with OMAM. The successful correction rate with OMAM was 100% (23/23) in group B and 82.6% (19/23) in group C. New reference values for corrected Doppler measurements independent of biometric variables were established. The SVIM with BSA successfully corrected none of the 23 parameters. In conclusion, different LV Doppler parameters allometrically correlated with one or more of the biometric variables. The novel OMAM developed in this study successfully corrected the effects of the physiological variances of most biometric variables on Doppler measurements in healthy and overweight subjects, and was found to be far superior to the SVIM. However, whether the OMAM equations can be applied to other ethnicities, obese subjects, and pathological conditions requires further investigation.