Flame atomic absorption spectroscopy (FAAS) is used to quantitatively determine trace metals in the test samples. The purpose of the present study is to treat some of the factors affecting AAS measurements. Some of the factors are under the control of the operator while other factors depend on the used equipment. In the present work equipment was used with low detection limits and high sensitivity but still some factors can't control. Using simulation methods to treat some of these measurements, obtain an estimate of the predicted concentration values for test samples. Two methods of simulation are investigated. The first simulation method is carried out by using line-of-best-fit through the experimental points using Excel programmed and using TREND function to estimate the unknown concentration directly from this calibration function. The second simulation method is based on the approximation programme which is built-in the concept of genetic programming (GP). This is carried out by the introduction of laboratory results of the known concentrations versus the absorption of known elements and finds the best curve of the approximate of each element. The results were rounded up to the laboratory results of most elements with low concentrations, the improvements of the methods were very satisfying and the results reached the imposed limits (A<0.05). Finally, a database was created for all results.