The double spike-standard addition (DSSA) is a powerful method proposed for isotope measurement of ultra-trace elements in samples. Here, we test the method's practicality and robustness for obtaining high-precision delta Cd-114/110 data in samples with complex matrices and ultra-trace cadmium (Cd) concentrations as low as 0.004 mu g g(-1). Through a DS data reduction routine and isotope binary mixing model, the delta Cd-114/110 values obtained for numeral geological and biological reference materials (GRMs/BRMs) agree well with certified or previous measurements, even for similar to 2.1 ng Cd of BCR-2. The overall precision of both single- and multi-standard addition depends on the sample fraction (f(spl)) owing to the error propagation. And 0.041 +/- 0.022 parts per thousand (2SD, n = 46) of precision for delta Cd-114/110 can be achieved when the sample fraction (f(spl)) is >= 20%, comparable to 0.056 +/- 0.039 parts per thousand (2SD, n = 23) obtained by the traditional DS method. However, it becomes greater than 0.110 parts per thousand when f(spl) is <20%, indicating that 20% (1/5) -50% (1/2) of f(spl) is the optimal mixing range to obtain high-precision data if the minimal sample sizes are required. These further confirm that accurate and precise Cd isotope ratios can be determined by DSSA. Correspondingly, the purification scheme can be simplified to a single column due to the added standard solution boosting analyte and diluting ratios of matrix/Cd. Animal organs, with 0.004-0.106 mu g g(-1) Cd, yield large variations of delta Cd-114/110 (-0.054 +/- 0.030 parts per thousand (2SD, n = 4) and 0.681 +/- 0.022 parts per thousand (2SD, n = 4) for ovine liver and kidney, respectively), suggesting that Cd isotopes can be fractionated significantly during biological metabolic processes and may be a potential use in medical diagnosis. Robust measurement of Cd isotope composition in ultra-trace Cd samples with complex matrices by DSSA broadens the scope of measurable samples by the traditional DS method, thus potentially opening a range of new opportunities in life, agricultural, environmental, and earth/planetary sciences.