Small nucleolar RNAs (snoRNAs) represent one of the largest groups of functional non-protein coding RNAs currently known in eukaryotic cells. The processing of intron encoded snoRNAs has been well documented; however, the transcriptional regulation of snoRNA genes is still poorly understood, most likely due to the lack of characterization of snoRNA promoters. Here we used a computational approach to predict core promoters for 131 human snoRNAs. Majority of putative snoRNA promoters are supported by DNase I hypersensitivities, RNA polymerase II ChIP-seq peaks, or CAGE tag clusters. Based on the genomic organizations of predicted human snoRNA promoters, we propose five transcriptional models of those snoRNA genes; we also found evidence that some intronic human snoRNAs might have their own promoters. The present study is the first in silico screening of human snoRNAs promoters; and we anticipate the data will facilitate further molecular characterization of transcriptional control of human snoRNA genes.