In 19 control patients and 10 patients with pernicious anemia (PA), the vitamin B12 (B12) absorption was determined simultaneously with whole‐body counting (FRB12) and with a double isotope technique employing incomplete stool collection (FAB12). The test dose was administered orally and consisted of 8 ml of a 10 ml solution containing 0.5 μg of 58Co‐B12 (approximately 0.4 μCi) and 2 mg of 51CrCI3 (approximately 20 μCi). Two ml of the solution were used as standard. In order to follow the passage of the inabsorbable tracer, 51CrCI3, the patients were given 25 radiopaque pellets and 4 carmine tablets (2 g) to swallow immediately after the test dose. Counting of a 3–4 ml specimen from one of the first two red stools was sufficient for calculating FAB12. The findings correlated closely with the FRB12 values (r=0.99, N=39, p<0.001). In the control subjects, the FAB12 averaged 74% (range 37–88). In the patients with PA given intrinsic factor (IF), the FAB12 averaged 40% (range 22–59). When IF was not given, FAB12 averaged 2% (range 0–9). The test, therefore, is a valid indicator of pernicious anemia. Reproducibility: For the mean value of 74%, the standard deviation (S.D.) was 5%, corresponding to a coefficient of variation (CV) of 7%. In the patients not given IF, S.D. was 1% and CV 50%, and in those given IF, S.D. was 8% and CV 20%. This shows the test to be as reliable as whole‐body counting. Carmine tablets proved to be a good indicator of isotope‐containing stool. The test is easy to perform and requires only a scintillation well‐counter. No co‐operation on the part of the patients is necessary. The test is also suitable for outpatients. Furthermore, it is independent of kidney function and flushing with B12 is avoided.