To accurately predict the main economic indices of high technology enterprises in China with nonlinear small sample characteristic, a new method for optimizing Nash nonlinear grey Bernoulli model (Nash NGBM(1,1)) is proposed in this study. Meanwhile, an optimized model is constructed to fully employ the predictive functions of original data information and solve optimum parameters. The process is indicated as follows: First, the merits and the disadvantages of the various approaches for ascertaining the initial conditions of the grey model are analyzed. Then, an optimized predictive function of the NGBM(1,1) is obtained by minimizing the error in the summed squares. Based on the function thus obtained, a nonlinear optimization model is developed to calculate the unknown parameters in the Nash NGBM(1,1). The results from a fluctuating sequence example and an actual case from the opto-electronics industry in Taiwan indicate that the optimized Nash NGBM(1,1) proposed in this paper gives a superior modeling performance. Finally, the main economic indices pertinent to high technology enterprises in China are forecasted using the optimized Nash NGBM(1,1) and relevant suggestions are made.