Real-time quantitative reverse transcription-PCR (qRT-PCR) has become one of the most commonly used methods for RNA quantification in recent years. To obtain reliable results with biological significance, it is important that qRT-PCR data are normalized with a proper internal control. In this study, 18 housekeeping genes were selected and evaluated for their potential as a suitable internal control for study of gene expression in the oomycete plant pathogen Phytophthora parasitica. Analysis of qRT-PCR data using the geNorm software indicated that, although commonly used as internal controls, beta-actin (ACT) and translation elongation factor 1alpha (eEF1A) might not be the best choice due to variable expression across different life stages of P. parasitica. Instead, other genes would serve as better controls, including ubiquitin-conjugating enzyme (Ubc), WS21, and beta-tubulin (Tub-b) for 'asexual stage,' Ubc and Tub-b for 'sexual reproduction,' while Ubc and WS21 for the stage of pathogenesis, because of their constant expression levels in each given subset of RNA samples. Although normalization with more than one gene would generate more reliable results, use of a single stably expressed gene as an internal control would suffice for accurate data normalization in some experiments.