1. Bioavailability of Fe from beef, spinach (Spinacea oleraceaL.) and their mixtures was studied using anaemic and non-anaemic rats by haemoglobin regeneration efficiency (HRE) and apparent Fe absorption in two trials.2. The initial haemoglobin levels of severely anaemic, mildly anaemic and non-anaemic rats were 63, 88 and 113 g/l, respectively. The Fe level in diets was about 30 mg/kg. All other nutrients equalled or exceeded the requirement of the growing rat.3. The spinach Fe was well utilized by the rats with average HRE of 0.41, 0.53 and 0.36, and apparent Fe absorptions averaging 0.48, 0.59 and 0.37 for the severely anaemic, mildly anaemic and non-anaemic animals respectively.4. Beef Fe was efficiently used by rats as reported by others. Average HRE were 0.42, 0.51 and 0.44, and average apparent Fe absorptions were 0.44, 0.47 and 0.46 for the severely anaemic, mildly anaemic and non-anaemic rats respectively.5. When the percentage of meat Fe was increased from 0 to 25, 50, 75 or 100 of the dietary Fe, HRE and apparent Fe absorption were not increased significantly. A meat enhancement effect on total Fe absorption, reported by others for non-haem-Fe, did not occur in the present experiment.6. Negative correlation coefficients between initial haemoglobin and HRE (r-0.79), and initial haemoglobin and apparent Fe absorption (r-0.73) were seen with the rats fed on dietary Fe from FeSO47H2O. This was not seen with the rats fed on dietary Fe from beef or spinach.7. The Fe absorption pattern for the different Fe sources is evidence for a third Fe pool, a pool made up of highly soluble inorganic Fe salt, in addition to haem-Fe and non-haem-Fe complex pools. FeSO47H2O is not in the same gastrointestinal pool as non-haem-Fe complex such as spinach Fe. A suggested mechanism of absorption is discussed.