In the present paper,uniform design U10(108) was used to optimize the condition of arsenic determination in vegetable samples by hydride generation-atomic fluorescence spectrometry.Mathematical model was established and regression analysis was done,and the optimized solutions to those equations were obtained by making use of the UD3.0 software.Combining the life-span of hollow cathode filament,noise of negative voltage and other factors,the optimal condition was obtained as follows:negative voltage was 280-360 V;lamp current was 50-70 mA;carrier gas flow rate was 500-700 mL·min-1;KBH4 concentration was 15.0-20.0 g·L^-1;HCL concentration was 0.6-1.2 mol·L^-1;sample size was 0.5-1.0 mL.Two samples of vegetable were analyzed under the optimized condition.The results showed that the relative standard deviation was less than 3.6%,and the recovery was within 94.1%-101.3%,with their detection limits of 0.42 μg·L^-1.In this paper,as an effective method of experiment design,uniform design was introduced to hydride generation-atomic fluorescence spectrometry analysis with multi-factors,which offered a good idea for the optimization of experiment conditions.