The paper aims to provide evidence on consistent condom use dynamics among sex workers in Central America between 1997 and 2000, and to examine the most important predictors of use behaviour important for policy and programme interventions in the region. Data on 3500 sex workers, 1500 from 1997 and 2000 from the year 2000, were analysed. The samples represented sex workers in low socioeconomic neighbourhoods who met their clients at known sex establishments or by the roadside. Sex workers were more likely to have used condoms consistently in 2000 than in 1997 (Odds Ratio (OR)=1·4, p≤0·05). Sex workers who discussed condoms with their partners or lovers used them consistently with all clients more than those who did not (OR=1·3, p≤0·10). Knowledge of condom advantages had a positive influence on consistent use. Sex workers who reported using condoms to prevent pregnancy or STDs used them consistently with all clients more than those who did not (OR=1·2, and 1·3 respectively, p≤0·10). The source of condoms is an important predictor of consistent condom use. Sex workers who bought condoms from health establishments or from brothels used them consistently more than those who did not (both OR=1·3, p≤0·10, and p≤0·05, respectively). Self-efficacy had a positive effect on consistent condom use. Sex workers who reported that they would use condoms even if clients offered to pay more for unprotected sex used them consistently with all clients more than those who did not (OR=1·8, p≤0·001). The findings suggest that having condom skill is positively related to condom use. Sex workers who had all four skills used condoms more consistently than those who had none (OR=1·6, p≤0·01). In order to increase consistent condom use and avert more incidences of HIV/AIDS and other STIs, programme interventions need to use the most efficient way to provide quality information, and provide repeated training on condom negotiation and use skills to sex workers.