We examined the effects of phosphorus (P) levels on photosynthetic and P/Fe traits of soybean under the stress of low Fe and their genotypic differences, to provide a theoretical basis for rational application of P and Fe fertilizer. Six P-efficient and six P-inefficient soybean varieties screened in the early stage were used as experimental materials. Four treatments of P:Fe ratio were set, including 0:30, 30:30, 150:30 and 300:30 (µmol·L-1). We measured chlorophyll fluorescence traits and P-Fe utilization efficiency in soybean. A stepwise regression equation was established with seed weight per plant. Pathway analysis was performed, with the response of P-efficient and P-inefficient soybean genotypes to different P:Fe treatments being comprehensively evaluated by factor scores. The results showed significant main and interactive effects of genotype and P:Fe on the relative electron transfer rate of photosystem Ⅱ (ETR) at beginning of flowering stage (R1), the proportion of the energy absorbed by photosystem Ⅱ dissipated into heat (NPQ) at R1 stage, and proportion of energy absorbed by photosystem Ⅱ devoted to the photochemical reaction (qL) at R1 stage. Results of canonical correlation analysis showed a negative correlation between P utilization efficiency of seed at full maturity stage (R8) and photosynthetic rate at R1 stage of P-efficient genotypes. Seed Fe utilization efficiency of P-inefficient genotypes at R8 stage was positively correlated with NPQ at R1 stage, but negatively correlated with qL at R1 stage. The actual photochemical efficiency of PSⅡ (ΦPSⅡ) at R1 stage was negatively correlated with P-efficient genotypes, but positively correlated with P-inefficient genotypes, which indicated that ΦPSⅡ at R1 stage was an important indicator for identifying soybean genotypes with different P efficiency under stress of low Fe. The comprehensive performance of P-efficient soybean genotypes decreased first and then increased with P level, while P-inefficient soybean genotypes increased first and then decreased. The inflection point of both genotypes appeared in P:Fe of 30:30. Thus, P:Fe ratio of 30:30 could be used as a threshold to identify soybean genotypes with different P efficiency under stress of low Fe. In conclusion, P fertilizer application should be equal to or greater than 1:1 (P:Fe) when planting P-efficient soybean genotypes in low Fe area, while P fertilizer application should not exceed 1:1 (P:Fe) when planting P-inefficient soybean genotypes.