Outdoor thermal environments strongly influence human energy balance and subjective comfort, requiring robust indices for reliable evaluation. The Predicted Mean Vote (PMV) model, though widely used in outdoor studies, was originally derived from controlled indoor experiments, raising concerns about its applicability in exterior contexts. This chapter examines the performance of PMV through field investigations conducted in Guangzhou during the summer. Results demonstrate a clear linear relationship between PMV and operative temperature, while the corresponding link between Thermal Sensation Votes (TSV) and operative temperature is less consistent. TSV increases steadily within the range of 28–36 °C, but tends to plateau at higher operative temperatures, remaining within 2.5–4.0 on the thermal sensation scale. In moderate conditions (28–34 °C), PMV aligns closely with TSV, whereas at elevated temperatures, PMV consistently overestimates thermal sensation. Based on statistical analysis, an adaptive correction model was developed, reducing discrepancies and offering improved reliability for outdoor thermal comfort evaluation. These findings provide both empirical validation and methodological refinement, contributing to the advancement of outdoor comfort assessment tools.