When a subject does non-contact healing, such as laying-on-of-hands, for sample pieces of cucumber "Shiro-ibo kyuuri", the intensity of biophotons from samples increases. To reconfirm the reliability of our measuring method for this phenomenon, we discussed the method from 3 view points: 1) validity of J value as an index of effects (J value is defined as the logarithm of the ratio of intensities of biophotons between experimental and control samples; 2) system biases; and 3) effects of heat or shadow (i.e. screening of light) of the hands. We re-analyzed previous data of laying-on-of-hands experiments and compared them with additional data. We concluded that the J value had sufficient validity as an index because its distribution was a normal distribution. Variation of intensities of biophotons of samples, differences of speed of drying or the rate of condensing of photons, drifts of temperature and background noise, effects of delayed luminescence, and difference of sensitivities of the measuring system were not considered to have any serious influence on the results. We found that in only the healing condition was the intensity of biophotons of experimental samples significantly larger than the intensity of controls (Wilcoxon signed-rank test, p=6.5×10^<-7>) and there was no significant difference in heat/shadow conditions. J value was 0.142 in the healing condition, but J values were nearly zero in the other conditions (ANOVA, p=0.03). We reconfirmed that healing effects are not the results of heat and shadow of the hands.