An experiment was conducted at Shangqiu Agro-Ecosystem Experimental Station in 2006–2008 to investigate the distribution and use efficiency of photosynthetically active radiation (PAR). The soil was classified as a Eutric Cambisol. Four treatments were compared: (i) sole crop maize (Zea mays L.) (SM); (ii) sole crop soybean [Glycine max (L.) Merr.] (SSB); (iii) three rows of soybean alternated with one row of maize (I1); and (iv) three rows of soybean alternated with two rows of maize (I2). Observed fraction of radiation intercepted (F), radiation use efficiency (RUE), and harvest index (HI) revealed that PAR transmittance at the bottom of the soybean canopy adjacent to the maize was higher than that farther from the maize. However, this difference decreased as the crops developed. The averaged extinction coefficient (K) and RUE were 0.51 and 2.82 g MJ–1 for I1 and 0.50 and 2.78 g MJ–1 for I2, respectively, compared with 0.46 and 3.18 g MJ–1 for sole crop maize and 0.59 and 1.55 g MJ–1 for sole crop soybean. The HI of I1 and I2 were 0.45 and 0.44 for maize and 0.35 and 0.36 for soybean, respectively. The HI of sole crop maize and sole crop soybean were 0.42 and 0.40, respectively. The mean land equivalent ratio (LER) for I1 and I2 was 1.65 and 1.71, respectively. Therefore, it can be concluded that intercropping usually had greater RUE than sole cropping, which may account for the yield advantage of intercropping.