Drug abuse is a major global public health problem. Fentanyl analogs and amphetamine-type stimulants （ATS） are widely abused worldwide. Hair analysis can trace drug use history over months. This technique has unique advantages in forensic toxicology. However， its application is limited by a lack of matrix-certified reference materials （CRMs）. At present， the existing CRMs are unable to simultaneously cover both fentanyl analogs and amphetamine-type stimulants. This study established a hair sample preparation method based on a water‑dimethyl sulfoxide‑hydrochloric acid soaking approach， targeting 15 analytes （nine fentanyl analogs and six ATS）. The soaking time was systematically investigated. Drug incorporation into hair reached a maximum at 24 days. A slight decrease was observed at 25 days. This decrease may result from desorption and solvent-induced structural changes in keratin. Therefore， 24 days was selected as the optimal soaking time. Preparation experiments demonstrated that 14 of the 15 drugs were successfully embedded into the hair matrix. This study also applied a dual-platform quantitative strategy for certification. The two platforms were liquid chromatography-tandem mass spectrometry （LC-MS/MS） and gas chromatography-mass spectrometry （GC-MS）. Cross-validation between the two platforms improved the reliability of certified values. Sample pretreatment conditions were optimized using Box-Behnken response surface methodology. Four factors were investigated： extraction temperature， extraction time， liquid-to-solid ratio， and hydrochloric acid concentration. The effects of the four factors on extraction efficiency followed a decreasing order： extraction temperature>extraction time>liquid-to-solid ratio>hydrochloric acid concentration. The optimal extraction conditions were as follows： extraction temperature 40 ℃， extraction time 50 min， liquid-solid ratio 100∶1 （mL/g）， and hydrochloric acid concentration 0.01 mol/L. Method validation was performed according to international guidelines. For LC-MS/MS， the limits of detection （LODs） and quantification （LOQs） were 0.05 pg/mg and 0.25 pg/mg， respectively. For GC-MS， the LODs and LOQs were 0.02 ng/mg and 0.08 ng/mg， respectively. Good linearities were obtained for all 15 drugs. Correlation coefficients （r） were greater than 0.999 for both platforms. Good linearities were obtained for all 15 analytes on both platforms with correlation coefficients >0.999 over the respective concentration ranges Intra-day precisions （n=6） were evaluated as relative standard deviations （RSDs）. Inter-day precision （n=30 over 5 days） was also calculated. For 14 drugs （except amphetamine）， RSD values ranged from 0.2% to 9.4%. All RSDs were below 10%， indicating good repeatability and intermediate precision. Stability of hair extracts was assessed at 1， 24， 48， 72， and 96 h at room temperature. The extracts were stable for at least 96 h when stored at room temperature， with RSDs below 6.0% for LC-MS/MS and below 10% for GC-MS. Recovery tests were performed at three quality control levels （low， medium， high）. For LC-MS/MS， average recoveries ranged from 84.46% to 116.4% （n=3）. For GC-MS， average recoveries ranged from 85.50% to 116.4% （n=3）. All RSDs were below 15%. Matrix effects （MEs） were evaluated by comparing matrix-matched standards with solvent standards. ME values between 80% and 120% （n=3） were observed for both platforms. These results indicate negligible ion suppression or enhancement. Homogeneity of the prepared hair reference material was assessed. The results show that the uniformity of the hair samples is good. The contents of the 14 drugs in the prepared hair samples were determined. For LC-MS/MS， the values ranged from 1.012 ng/mg to 7.830 ng/mg. For GC-MS， the values ranged from 1.087 ng/mg to 7.712 ng/mg. Amphetamine was not detected by either platform， likely due to its high polarity and weak binding to keratin. Most drugs showed good agreement between the two platforms. Relative deviations were below 10% for most analytes. However， methamphetamine showed a deviation of -12%. This may be due to matrix effects in LC-MS/MS. The GC-MS results for methamphetamine were more robust. The dual-platform approach allowed cross-validation and enhanced the reliability of the certified values. In conclusion， a dual-platform method based on LC-MS/MS and GC-MS was successfully applied to certify a new hair matrix reference material. The cross-validation strategy enhances the accuracy and traceability of the certified values. This work provides a valuable tool for quality control in forensic toxicology laboratories. The prepared reference material and validated method will facilitate inter-laboratory comparisons. They will also support the monitoring of drug abuse in forensic and clinical settings.