An increasing focus on environmental sustainability has led to an ongoing global effort to better understand and engineer platform organisms that economically convert CO2 and other waste gases into useful biofuels and chemicals. Using stoichiometric and energetic analyses, we assess the efficiencies of four non-photosynthetic CO2 fixation pathways of practical importance, with a focus on engineered strains. The analysis compares the pathways based on their ATP and H2/electron requirements, the number of enzymes required, and the production of three model chemicals: ethanol, acetate, and butanol. Our analysis shows that the Wood–Ljungdahl (WL) pathway is the most efficient, based on the most expensive substrate (i.e. H2 or electrons), for the production of acetate and ethanol. Chemoautotrophically, butanol formation is an ATP-limited process, making anaerobic production from the WL pathway inefficient; however, higher potential butanol titers are predicted for the aerobic pathways. Mixotrophic growth, in which organic substrates are fed alongside H2/CO2, alleviates the ATP limitations and thus improves the yields for both aerobic and anaerobic butanol production. We also calculate maximal yields, both chemoautotrophic and mixotrophic, based on the WL pathway for two other important molecules: 2,3-butanediol and butyrate.