The photosynthetic models now mainly consider the effects of environmental factors on plant photosynthesis. These models can only predict photosynthetic rate of leaves with similar physiological conditions. In order to meet the needs of modeling the models for leaf photosynthetic rate prediction in different growth states, a method for constructing a multienvironmental factors photosynthetic rate prediction model incorporating dark fluorescence parameters Fv/Fm was proposed. Taking eggplant leaves of different growth states as samples, the Fv/Fm were obtained, and the photosynthetic rates were measured at different temperatures, CO2 concentrations and light intensities to construct a set of modeling samples. Then a unified prediction model of photosynthetic rate was established by using genetic support vector regression. The determinant coefficient of the model was 0.8895, and the root mean square error was 3.2679μmol/(m2〖DK〗·s). The results of XOR checkout showed that the accuracy of the model was improved remarkably by fusing the Fv/Fm. The fitting slope between the predicted and measured photosynthetic rates was 0.9046, the intercept was 0.3641, which showed that the model could predict an exact photosynthetic rate of leaves with different physiological conditions by leading in Fv/Fm.