The Technology Acceptance Model (TAM) and Theory of Planned Behavior (TPB) are widely used in the information systems (IS) literature to explain and predict user technology acceptance behavior. Although these models have been utilized separately in the earlier studies, to increase the richness and predictive power of their studies, scholars have combined both theories in more recent studies. However, few studies have compared these models directly using the same empirical data. The justification for using various hybrid models is mostly anecdotal and convenient rather than based on strong theoretical arguments. Thus the question remains: when studying user technology acceptance behavior, which model offers the best fit to empirical evidence? This study compares four different models (TAM, TPB and two combined TAM-TPB) in terms of their predictive power and model fit to empirical data using meta-analysis methodology. The data for this study is gathered from metaanalytic calculations of 32 studies found in the technology acceptance literature. Our findings demonstrate that when used separately TAM and TPB models offers not only theoretical parsimony and clarity, but also better fit with empirical data than the two combined TAM-TPB models. Implications of our findings are discussed and future research directions are presented.