The determination of morphine and its isobaric metabolites morphine-3-beta-d-glucuronide (M3G) and morphine-6-beta-d-glucuronide (M6G) is useful for therapeutic drug monitoring and forensic identification of drug use. In particular, capillary electrophoresis with mass spectrometry (CE-MS) represents an attractive tool for opioid analysis. Whereas volatile background electrolytes in CE often improve electrospray ionization for coupled MS detection, such electrolytes may reduce CE separation efficiency and resolution. To better understand the effects of background electrolyte (BGE) composition on separation efficiency and detection sensitivity, this work compares and contrasts method development for both volatile (ammonium formate and acetate) and nonvolatile (ammonium phosphate and borate) buffers. Peak efficiencies and migration times for morphine and morphine metabolites were optimal with a 25mM ammonium borate buffer (pH=9.5) although greater sensitivities were achieved in the ammonium formate buffer. Optimized CE methods allowed for the resolution of the isobaric morphine metabolites prior to high mass accuracy, electrospray ionization quadrupole time-of-flight (ESI-QTOF) MS detection applicable to the analysis of urine samples in under seven minutes. Urine sample preparation required only a 10-fold dilution with BGE prior to analysis. Limits of detection (LOD) in normal human urine were found to be 1.0μg/mL for morphine and 2.5μg/mL for each of M3G and M6G by CE-ESI-QTOF-MS. These LODs were comparable to those for CE-UV analysis of opioid standards in buffer, whereas CE-ESI-QTOF-MS analysis of opioid standards in buffer yielded LODs an order of magnitude lower. Patient urine samples (N=12) were analyzed by this new CE-ESI-QTOF-MS method and no significant difference in total morphine content relative to prior liquid chromatography-mass spectrometry (LC-MS) results was found as per a paired-t test at the 99% confidence level. Whereas the LC-MS method applied to these samples determined only total morphine content, this new CE-ESI-QTOF-MS method allowed for species differentiation in addition to total morphine determination. By this method, it was found that M3G and M6G metabolites were present in a 5:1 concentration ratio, on average, in patient samples. Therefore, the CE-ESI-QTOF-MS method not only allows for total morphine concentration determination comparable to established LC-MS methods, but also allows for differentiation between morphine and its trace glucuronides, yielding additional biochemical information about drug metabolism.