In the present paper, the slope gradient, aspect, detection zenith angle and plant types were analyzed. In order to strengthen the theoretical discussion, the research was under laboratory condition, and modeled uniform slope for slope plant. Through experiments we found that these factors indeed have influence on plant hyperspectral remote sensing. When choosing slope gradient as the variate, the blade reflection first increases and then decreases as the slope gradient changes from 0° to 36°; When keeping other factors constant, and only detection zenith angle increasing from 0° to 60°, the spectral characteristic of slope plants do not change significantly in visible light band, but decreases gradually in near infrared band; With only slope aspect changing, when the dome meets the light direction, the blade reflectance gets maximum, and when the dome meets the backlit direction, the blade reflectance gets minimum, furthermore, setting the line of vertical intersection of incidence plane and the dome as an axis, the reflectance on the axis’s both sides shows symmetric distribution; In addition, spectral curves of different plant types have a lot differences between each other, which means that the plant types also affect hyperspectral remote sensing results of slope plants. This research breaks through the limitations of the traditional vertical remote sensing data collection and uses the multi-angle and hyperspectral information to analyze spectral characteristics of slope plants. So this research has theoretical significance to the development of quantitative remote sensing, and has application value to the plant remote sensing monitoring.