We measure power electricfield intensity at a substation in southwestern P. R. China at various humidity levels and temperatures. Measurement data and a back propagation neural network are used to construct a model of electrical field intensity responses to fluctuations in humidity and ambient temperatures. The maximum error is 0.047 5. The results show that the power electricfield intensity increases as humidity rises when the temperature is constant; when humidity is a constant, temperature changes do not affect the regularity of the electricfield intensity in a consistent manner regularly.