Introduction Road traffic accidents are among the leading causes of death in Uganda. Taxi operators are at a high risk of RTA and can potentially be first responders. This study, aimed to assess knowledge, attitude, and practice of first aid among taxi operators in the new taxi park, Kampala Uganda. Methods A descriptive cross-sectional study was conducted in 2021 among taxi drivers and conductors in the New Taxi Park, Kampala City, Uganda. A semi-structured questionnaire was used to collect quantitative data from taxi operators after informed consent. Chi-square or Fisher's exact test and logistic regression were performed in STATA 16 to assess the association between first aid knowledge and demographics. P < 0.05 was statistically significant. Results A total of 345 participants, majority males ( n = 338, 98%) aged between 18 and 45 years (76.5%), were recruited. Although 97.7% ( n = 337) had heard about first aid, only 19.4% ( n = 67) had prior first aid training. Overall mean knowledge score was 40.1% (SD = 14.5%), with 29.9% ( n = 103) having good knowledge (≥50%). Participants who had witnessed more than five accidents (aOR = 2.9, 95% CI = 1.7–4.8, p < 0.001), those with first aid kits (aOR = 1.7, 95% CI = 1.0–3.0, p = 0.38) were more likely to have good knowledge while those below post-secondary education level i.e., Primary (AOR = 0.2, 95% CI = 0.1–0.5, p ≤0.001) and secondary (aOR = 0.2, 95% CI = 0.1–0.6, p = 0.001), were less likely to have good knowledge. About 97% and 93% perceived first aid as important and were willing to undergo training, respectively; however, only 69% were willing to give first aid. Only 181(52.5%) had ever attended to accident victims. Conclusion Majority of taxi operators had poor first aid knowledge. Factors associated with good knowledge included level of education, number of accidents witnessed, having first aid kits. Although their attitudes were favorable, practice was poor. Comprehensive training and refresher courses can help increase first aid knowledge, and improving practice.