Background Lab-based surveillance (LBS) of antimicrobial resistance (AMR) is not systematically implemented in Uganda. We aimed to identify the gaps in establishing regular LBS of AMR in Uganda. Methods This was a retrospective records review. It was done at Mulago Hospital (MH) Microbiology Lab (MHL). It analyzed lab records on bacteria and their antimicrobial susceptibility profiles (ASPs) over 6 months. Since MH is the national referral hospital, we hypothesized that (1) MHL is the best resourced and that any limitations seen here are amplified in labs at regional referral hospitals (RRHs) and health centers (HCs); (2) data from MHL on LBS mirrors that from labs at RRHs and HCs; (3) the state of records from MHL show lab performance and the presence or absence of standard operating procedures (SOPs), as would be the case at RRHs and HCs. Results The lab got 1760 samples over the six months. The most common sample was urine (659, 37.4%). From the 1760 samples, data on 478 bacterial isolates were available. Urine gave the most isolates (159, 33.3%). Most of the isolates were gram-negative (267, 55.9%). Escherichia coli (100, 37%) was the most common gram-negative (of the Enterobacteriaceae). Pseudomonas aeruginosa (17, 6%) was the most common gram-negative (of the non-Enterobacteriaceae). The gram-negative bacteria were highly resistant to β-lactams. These were highly sensitive to carbapenems. The Staphylococcus aureus were highly resistant to β-lactams. The gram-positive bacteria were highly sensitive to vancomycin. ASPs for all bacterial categories were incomplete. Conclusions The findings from MHL suggest that there is a need to improve antimicrobial susceptibility testing capacity. They also suggest that the situation at RRHs and HCs around the country could be worse. So, there is a need to extend the political will, which already exists, into operational and implementation action.