Plastic nets with opaque threads are frequently used for shading agricultural structures under high solar radiation conditions. A parameter that is often used to define a net is the net porosity (Pi). Value of Pi is usually estimated by one of three methods: image processing, direct beam transmittance, or solar radiation balance (hereafter radiation balance). Image processing is a rather slow process because it requires scanning the net sample at high resolution. The direct beam transmittance and radiation balance methods greatly overestimate Pi because some of the solar radiation incident on the thread surfaces is forward scattered and add a considerable amount of radiation to that transmitted from the net pores directly. In this study, the radiation balance method was modified to estimate Pi precisely. The amount of solar radiation scattered forward on the thread surfaces was estimated separately. Thus, the un-scattered solar radiation transmitted from the net pores directly, which describes the net porosity, Pi could be estimated. This method, in addition to the image processing and the direct beam transmittance methods were used to estimate Pi for different types of nets that are commonly used for shading structures in summer. Values of Pi estimated by using the proposed method were in good accordance with those measured by the image processing method at a resolution of 4800 dpi. The direct beam transmittance and the radiation balance methods resulted in overestimation errors in the values of Pi. This error strongly depends on the color of the net. The estimated errors were +14% for a green net and +37% for a white net when using the radiation balance method, and were +16% and +38%, respectively, when using the direct beam transmittance method. In the image processing method, a resolution of 2400 dpi is sufficient to estimate Pi precisely and the higher resolutions showed no significant effect on the value of Pi. (C) 2010 Elsevier Ltd. All rights reserved.