A slurry sampling method is proposed for the determination of iron and zinc in baby food by flame atomic absorption spectrometry without sample digestion prior to analysis. The effect of slurry concentration (the ratio of solid sample to total slurry volume), different acids at various concentrations as diluent and the addition of dispersant on accuracy and precision were investigated. The samples were dried at 105 degrees C overnight and ground using an agate mortar. To obtain quantitative recoveries, 500 mg of sample was slurried in 20 ml of 0.05% Triton X-114 containing 0.1 M HNO3, homogenised using a high-performance overhead disperser at 15,000 rpm for 5 min and directly aspirated into the flame. The accuracy of the method was tested by determination of analytes in various certified reference materials. The limits of detection of the method (N = 10; 3 sigma) for iron and zinc were 5.5 and 3.4 mu g g(-1), respectively, using a very dilute slurry of baby food, which gave a very low background signal. Finally, the proposed method was applied to the determination of iron and zinc in different baby food samples obtained from markets in Turkey. The range of iron and zinc content for the samples were 33-76 and 15 73 mu g g(-1), respectively.