Abstract: A systematic understanding of the multiscale statistical characteristics of water-vapor fluctuations under dust-laden conditions and their relationship with turbulent transport is still lacking. This study is based on synchronous observations in the atmospheric surface layer at Qingtu Lake. Eight dust-laden datasets and twelve clean-air datasets obtained under near-neutral conditions are selected for comparison. Power spectra and autocorrelation functions are used to characterize the distribution of specific humidity across scales and its streamwise correlation. Wavelet flatness is used to assess intermittency. Linear coherence spectra and cospectra are used to examine velocity-specific-humidity coupling and scale-dependent flux contributions, while quadrant analysis is used to quantify the contributions of different turbulent events. The results show that, under dust-laden conditions, the spectral peak of specific humidity shifts toward lower wavenumbers and its autocorrelation decays more slowly, indicating that the relative distribution of specific-humidity variance shifts toward larger scales. The mean wavelet flatness of specific humidity is generally lower than that under clean-air conditions, suggesting a tendency toward reduced intermittency. Within the wavenumber range resolved by the observations, the linear coherence between velocity and specific humidity is enhanced at relatively high wavenumbers in the dust-laden samples. The premultiplied streamwise-velocity-specific-humidity and vertical-velocity-specific-humidity cospectra are dominated by negative and positive contributions, respectively. Ejection and sweep events make the principal contributions to the fluxes. Together, these events account for approximately 75%-95% and 80%-98% of the sum of the absolute values of the net contributions from all four quadrants to the streamwise and vertical water-vapor fluxes, respectively. In the dust-laden samples analyzed here, ejection and sweep events contribute predominantly negatively to the streamwise water-vapor flux and positively to the vertical flux, indicating different transport roles in the two directions. The difference between the scales associated with the distribution of specific-humidity variance and those associated with flux contributions indicates that water-vapor transport should be interpreted using the joint statistics of velocity and specific humidity. These results provide observational evidence for understanding turbulent water-vapor transport in dust-laden environments.