Skill-biased technical change occupied empirical economists for much of the 1990s. The empirical literature firmly established a positive correlation between technology indicators and demand shifts. In the minds of many, that has established a causal relationship. This leap of faith, however, is at odds with Hicks's conventional wisdom that endogenous technological change will be biased toward using cheap and abundant resources. In addition, if the rate of technical change is considered to be endogenous, the assumption of an exogenous bias toward skilled labor should at least be questioned. Two hypotheses explaining endogenous skill bias in technical change have been suggested in the theoretical literature: the acceleration effect and the market size effect. In this paper these are studied in a single endogenous-growth model to derive the sufficient and necessary conditions for both hypotheses. After confronting these conditions with the evidence, the paper concludes that it strongly favors the acceleration hypothesis.