Quality of life (QOL) is considered as a global, yet unidimensional, subjective assessment of one's satisfaction with life. We examine the construct validity of the available indicators of global QOL by constructing a causal model in which QOL is viewed as causally responding to several dimensions of perceived health. Global QOL is measured with five indicators derived from the literature, and perceived health is measured by eight dimensions of the SF-36 Health Survey. The structural equation model describes the QOL and perceived health assessments of 306 coronary artery bypass graft (CABG) patients and provides a stringent test of the unidimensionality of the global QOL indicators. The evidence regarding QOL is worrisome, but not devastating. The acceptable model fit was achieved by introducing direct effects leading from the General Health Perceptions concept to two of the five indicators of global QOL. This suggests that there are some mechanisms that influence these indicators without first altering QOL, or that these indicators are not responding in unison to QOL. Furthermore, the Faces QOL indicator required measurement error correlations to two of the perceived health indicators. This also suggests that there are mechanisms beyond pure QOL that act upon this indicator. But the problematic effects and error correlations are modest, so the evidence against these indicators is not particularly strong. Overall, the Self-Anchoring Striving Scale (SASS) indicator and the Life-as-a-Whole indicator (from Multiple Discrepancies Theory) seem to be the cleanest indicators of global QOL. General health perceptions and perceived mental health both influenced global QOL, while the other six health perception concepts had no consistent impact on global QOL.