The staining of nucleic acids with fluorescent dyes is one of the most fundamental technologies in relevant areas of science. For reliable and quantitative analysis, the staining efficiency of the dyes should not be very dependent on the sequences of the specimens. However, this assumption has not necessarily been confirmed by experimental results, especially in the staining of ssDNA (and RNA). In this study, we found that both SYBR Green II and SYBR Gold did not stain either homopyrimidines or ssDNA composed of only adenine (A) and cytosine (C). However, these two dyes emit strong fluorescence when the ssDNA contains both guanine (G) and C (and/or both A and thymine (T)) and form potential Watson-Crick base pairs. Interestingly, SYBR Gold, but not SYBR Green II, strongly stains ssDNA consisting of G and A (or G and T). Additionally, we found that the secondary structure of ssDNA may play an important role in DNA staining. To obtain reliable results for practical applications, sufficient care must be paid to the composition and sequence of ssDNA.