Introduction Ethyl glucuronide (EtG) in hair is currently used as a long-term alcohol marker, and methods with low LOQs are required to differentiate teetotalers from moderate drinkers according to the cut-offs (i.e., 7pg/mg hair). The determination of EtG may be performed by either LC-MS/MS or GC-(MS/)MS. The use of negative ion chemical ionization (NICI) improves the analytical sensitivity of GCMS, whereas similar LOQs were observed for GC-MS and GC-MS/MS (C.L. Crunelle et al., Drug Alcohol Depend. 2014;134:1-11). However, the direct comparison of GC-MS and GC-MS/MS in NICI has never been investigated. This study compares GC-MS and GCMS/ MS in NICI mode for EtG determination in hair in order to assess the additional qualitative and quantitative advantages of GCMS/ MS. Methods Hair segments from 64 volunteers (teetotalers, moderate drinkers and excessive drinkers) were analyzed by GC-MS and GCMS/ MS operated in NICI mode on the same instrument. Samples were processed according to I. Kerekes et al. (Ther. Drug Monit. 2013;35:527-529) using a derivatization with heptafluorobutyric anhydride. Results The LOD and LOQ obtained using GC-MS/MS were 0.005pg/mg and 0.017pg/mg respectively, compared to a LOD of 0.022pg/mg and LOQ of 0.075pg/mg with GC-MS. R2 of calibration curves of both methods were above 0.998. Furthermore, lower background noise was observed using GC-MS/MS. The obtained EtG concentrations from the hair samples were similar (mean CV=1.01 %). The differences between both methods were larger for lower EtG concentrations (mean CV=1.67 %) than for higher EtG concentrations (mean CV=0.64 %). This suggests a higher selectivity of GC-MS/MS at lower concentrations. Conclusions By using GC-MS/MS, a higher analytical selectivity and an improved signal to noise ratio, can be achieved. However, using GC-MS/MS instead of GC-MS would not change interpretation of the samples’ EtG concentration regarding the differentiation between teetotalers, moderate drinkers and excessive drinkers.