A flow-batch manifold coupled to a flame atomic absorption spectrometer was evaluated to assess the iron content by the internal standard method in hydrated ethanol used as fuel in automotive industry. For this assessment official methods require calibration procedures with matrix matching, making it difficult to obtain accurate results for samples adulterated by the addition of water. Nickel was selected as the internal standard since it is usually absent in samples and because it requires similar conditions of atomization. After procedure optimization, which requires about 4.25mL of sample and standard per measurement, it was possible to get linear analytical response for iron concentrations between 0.12 and 1.40mgL(-1) and a detection limit of 0.04mgL(-1). Eighteen samples were collected randomly from fuel stations in Pernambuco (Brazil) and iron concentration was determined using the proposed procedure. Comparison of results obtained (0.20-1.50mgL(-1)) showed a mean standard error of 3.9%, with 3.8% and 2.3% calculated for the mean variation coefficients of the proposed method and the reference procedure, respectively. For adulterated samples (0.12-0.64mgL(-1)), the mean standard error was 4.8% when compared with the standard addition method. These results allowed concluding that the proposed procedure is adequate to accomplish the determination of iron in ethanol fuel in a large scale basis with a sampling rate of about 10h(-1).