We initiated a comparison of Candida albicans stationary-phase gene expression with other growth states. The widely used hot acid phenol method for RNA extraction did not extract rRNA from late stationary-phase cells. The RNA from growing yeast cells, hyphae and biofilm was biased towards small-sized RNA. The 2 : 1 ratio between the two large rRNA bands was rarely obtained. Real-time reverse transcriptase PCR was used to determine mRNA extraction by several methods for OXR1, IRA2, RAD50, PNC1 and CHS2, which have 300 bp-8 kb coding regions, and ACT1, EFB1 and TDH3, sometimes used as internal standards. Only smaller-sized cDNA species were amplified from some extracts. Crushing cells with glass beads in liquid nitrogen before RNA extraction by the hot phenol method (CGB) yielded an unbiased distribution for rRNA and mRNA as verified by real-time reverse transcriptase PCR. With the CGB method, the large mRNA species, RAD50, IRA2 and OXR1, were present throughout the stationary phase, whereas the CSH2 transcript increased. The ACT1, EFB1 and TDH3 transcripts decreased in the stationary phase, making them unsuitable for standardization. The CGB method yielded high-quality RNA with the various growth conditions and permitted the comparison of stationary-phase transcripts with those obtained under other conditions.