OBJECTIVES: The purpose of this study was to determine the pathologic basis of Q-wave (QW) and non-Q-wave (NQW) myocardial infarction (MI). BACKGROUND: The QW/NQW distinction remains in wide clinical use but the meaning of the difference remains controversial. We hypothesized that measurement of total MI size and transmural extent by late gadolinium enhancement cardiovascular magnetic resonance (CMR) would identify the pathologic basis of QWs. METHODS: A total of 100 consecutive patients with documented previous MI had electrocardiogram and CMR on the same day. Patients with acute MI within seven days were excluded. Left ventricular function and the size and transmural extent of MI were quantified in the three major arterial territories and correlated with the presence of QW. RESULTS: Subendocardial MI showed QW in 28%. Transmural MI showed NQW in 29%. Of all MIs, 48% were at some point transmural, and 99% of these were at some point non-transmural. As MI size and number of transmural segments increased, the probability of QW increased (anterior: total size chi-square = 53, p < 0.0001, transmural extent chi-square = 36, p < 0.0001; inferior: total size chi-square = 16, p = 0.001, transmural extent chi-square = 10, p = 0.001). These findings did not hold for lateral MI. In a multivariate model, the transmural extent of MI was not an independent predictor of QW when total size of MI was removed. The QW/NQW classification was a good test for size of MI (area under receiver operating characteristic curve: anterior 0.90, inferior 0.77). CONCLUSIONS: The QW/NQW distinction is useful, but it is determined by the total size rather than transmural extent of underlying MI.