Abstract: In most conventional studies, design and parameter optimization of atomic absorption spectrometer(AAS) is usually carried out using a single-varialble-at-a-time (SVAT, the most common practice holding all other variable constants) method, but the method is time-consuming, cannot identify the interactions between different variables, and is inefficient to predict the true optimized parameters of the operating conditions. The central composite design(CCD) is one of the methods based on the response surface methodology(RSM). CCD used in this study can overcome the above mentioned shortcomings. Copper is used as the target compound for the test, and the Cu concentrations were analyzed by AAS with different operating parameters of such as acetylene flow, lamp current and sample inject rate. The results showed, that the optimized conditions were acetylene flow at 3.00 L/min, lamp current at 10 mA or sample inject at 7 mL/min by the single-varialble-at-a-time. Three experimental variables including acetylene flow, lamp current and sample inject were selected in the multivariable experimental design. The optimized conditions were:acetylene flow at 2.25 L/min, lamp current at 10 mA and sample inject 7 mL/min.