The bioavailability of iron from ferrous sulfate sources that had been stored for unknown periods was evaluated by using iron-deficient rats in experiment 1. The efficiency of converting dietary iron from these sources into hemoglobin varied from 60 to 84%. This variability was not explained on the basis of percent of iron in the ferrous state nor on the percent of iron in the source. The one fresh source used was reevaluated 3 months later, and its bioavailability had been reduced from 84 to 65%. The bioavailability of six fresh ferrous sulfate samples varied from 50 to 59% in experiment 2. It was concluded that, whenever ferrous sulfate is to be used as a reference source, fresh salt should be obtained to reduce variability among experiments.