The ability to fluently read music notation is highly variable among music learners (Gudmundsdottir, 2010; Kuhlman, 2005; Mills & McPherson, 2015). The abstract and culturally mediated nature of the psychological association, or cross-modal correspondence, between auditory pitch and visuospatial height that underlies the Western staff notation system (e.g. Dolscheid et al., 2015; Elkoshi, 2019; Flowers & Costa-Giomi, 1991; Lee, 2013; Speed et al., 2021; Starr & Srinivasan, 2018) means that music learners may vary in how intuitive they find reading pitch notation to be.To determine the potential impact of individually variable strengths of cross-modal correspondence between pitch and height on music reading skills, we first must be able to measure the presence and nature of this phenomenon in a way that captures its complexity. The purpose of this study was to determine the validity and reliability of the Pitch-Height Stroop Test (PHST), a novel measure of cross-modal correspondence between pitch and both perceptual and linguistic aspects of spatial height. Research questions included: 1.) Is the PHST a valid and reliable measurement tool? 2.) To what extent are the magnitudes of various cross-modal correspondences between pitch and linguistic or perceptual representations of height linked? 3.) Are the magnitudes of cross-modal correspondences between pitch and height predictive of sight-singing fluency? If so, which ones? 4.) Are these magnitudes associated with extent of music training?English-speaking adult singers (n = 50) completed the PHST along with a demographic and musical background questionnaire, the Profile of Music Perception Skills pitch and melody subtests (Zentner & Strauss, 2017), a flanker task (Eriksen & Eriksen, 1974), and the Vocal Sight-Reading Inventory (Henry, 1999). Results showed that the PHST was a largely valid and reliable measure. Magnitudes of pitch-height cross-modal correspondence appeared to reflect differences between unisensory and cross-sensory processing and showed moderate correlation between its linguistic and perceptual forms. Regression analysis indicated that perceptual pitch-height cross-modal correspondence positively predicted 10% of sight-singing fluency variance, whereas its linguistic form did not significantly contribute. Correlations between pitch-height cross-modal correspondence and musical background were limited. These results indicate that perceptual pitch-height cross-modal correspondence may be an important factor when learning to read music, and that the ability to verbally identify pitches as “high” versus “low” may be helpful but not sufficient to build the auditory-visual associations necessary for fluent music reading. The PHST may be a useful tool for future investigations of whether it is possible to increase this phenomenon through intentional instruction.