Separation and detection of native anhydrous carbohydrates derived from the combustion of biomass using an electrophoretic microchip with pulsed amperometric detection (PAD) is described. Levoglucosan represents the largest single component of the water extractable organics in smoke particles and can be used to trace forest fires or discriminate urban air pollution sources. Detection of levoglucosan and other sugar anhydrides in both source and ambient aerosol samples is typically performed by gas chromatographic (GC) separation with mass spectrometric (MS) detection. This method is cost, time, and labor intensive, typically involving a multistep solvent extraction, chemical derivatization, and finally analysis by GC/MS. However, it provides a rich wealth of chemical information as the result of the combination of a separation method and MS and exhibits good sensitivity. In contrast, microchip capillary electrophoresis offers the possibility of performing simpler, less expensive, and faster analysis. In addition, integrated devices can be fabricated and incorporated with an aerosol collection system to perform semicontinuous, onsite analysis. In the present report, the effect of the separation potential, buffer pH and composition, injection time, and pulsed amperometric detection parameters were studied in an effort to optimize both the separation and detection of anhydrous sugars. Using the optimized conditions, the analysis can be performed in less than a minute, with detection limits ranging from 22 fmol (16.7 microM) for levoglucosan to 336 fmol (258.7 microM) for galactosan. To demonstrate the capabilities of the device, a comparison was made between GC/MS and microchip electrophoresis using an aerosol source sample generated in a wood-burning chamber. A second example utilizing an ambient aerosol sample illustrates a matrix interference necessitating additional method development for application to samples not dominated by wood smoke.