Objectives We compared the prevalence of inconsistent condom use during commercial sex between female sex workers (FSWs) who did or did not inject drugs (FSW-IDUs and FSW-NIDUs) and investigated factors associated with this inconsistent use within these two groups. Methods: Some 158 FSW-NIDUs recruited from sex work venues and 218 FSW-IDUs recruited via the snowball sampling method were interviewed anonymously. Results: Only 16.5% of the FSW-IDUs and 51.3% of the FSW-NIDUs had used condoms consistently during commercial sex in the last month (odds ratio (OR) = 0.19). Factors significantly associated with inconsistent condom use in both groups included: behavioural intention for condom use (adjusted odds ratio (AOR) = 0.05 and 0.13), condom unavailability (AOR = 4.77 and 5.33), a perceived need to engage in unprotected sex if the client paid more (AOR = 8.74 and 10.84) or insisted on demanding unprotected sex (AOR = 19.78 and 7.59), and submissive gender power (AOR = 11.65 and 2.58). One factor, perceived susceptibility (AOR = 2.64), was significant only among FSW-NIDUs, whereas perceived efficacy of condom use in preventing HIV transmission (AOR = 0.08), perceptions that peer FSWs would not use condoms with clients (AOR = 2.23), self-hatred (AOR = 2.25) and lack of social support (AOR = 2.93) were significant only among FSW-IDUs. Injecting with used syringes was also associated with inconsistent condom use among FSW-IDUs (AOR = 4.64). Conclusions: FSW-IDUs were more likely than FSW-NIDUs to possess the cognitive and psychosocial conditions associated with unprotected commercial sex. Interventions need to take these differences into account.