Background Lymphogranuloma venereum (LGV) is a rare sexually transmitted infection (STI), a biovar of Chlamydia trachomatis , which re-emerged in 2003 and mainly affects HIV-positive men who have sex with men (MSM). A diagnostic service at the National Reference Laboratory was established in response to the LGV outbreak in the UK, which has been the largest documented in Europe. Surveillance protocol operates through testing patients with symptoms indicative of the infection and their sexual contacts. Additionally, a voluntary enhanced surveillance system was ongoing in 2004–10 in the UK. We assess the utility of the novel surveillance system in light of its strengths and limitations. Methods We described data on samples tested for LGV in 2004–10 in the diagnostic service. These data were analysed together with the LGV enhanced surveillance dataset and we undertook a cross-sectional analysis of clinical and behavioural characteristics between HIV-positive and HIV-negative or unknown cases diagnosed in MSM in 2004–10. We used multivariable logistic regression models with generalised estimating equations to control for repeat infections. The final model was adjusted for clinical and behavioural variables chosen through selection of variables based on a-priori hypotheses and statistical associations. Findings Overall, 11 196 C trachomatis -positive samples were tested for LGV, most of which were from men (10 035 of 11 196 [89·6%]), and most of these were from rectal sites (9138 of 10 035 [91·1%]); 1523 (15·2%) of 10 035 samples in men and four (0·8%) of 531 samples in women were confirmed as LGV. Of those confirmed to have the disease, 1370 (86·7%) of 1581 had a LGV enhanced surveillance form. The enhanced surveillance included 1342 episodes of infection in 1281 MSM and most were known to be HIV positive (1028 of 1281 [80·2%]). In comparison with HIV-negative and unknown cases, HIV-positive men were more likely to report unprotected receptive anal intercourse in the preceding 3 months (adjusted odds ratio [aOR] 2·7, 95% CI 1·3–5·8). However, HIV-positive men were also more likely to report a shorter duration of symptoms before clinic presentation (aOR 0·5, 95% CI 0·3–0·8 for reporting more than a week compared with a week or less), and the recorded information on the clinical variables was more complete for HIV-positive MSM. Interpretation There has been good coverage with the enhanced surveillance data, probably because of its link to a centralised diagnostic capacity. The surveillance was able to focus on the subpopulation at greater risk of infection because risk behaviour was more common in HIV-positive than in HIV-negative or unknown LGV cases. Furthermore, the surveillance has shown a stable epidemiological profile of cases over time, and LGV seems to remain confined to a small subpopulation of HIV-positive MSM. However, diagnostic bias was suggested in favour of HIV-positive cases with LGV who presented with a shorter duration of symptoms than HIV-negative or unknown cases. This factor is probably a result of the protocol used for testing together with clinical requests for tests. Surveillance for LGV may be systematically missing asymptomatic cases as well as there being a delay in recognising cases in HIV-negative men. Control of the disease needs targeted prevention messages that are suitable for the key population affected, including outreach and direct contact to key venues and networks via the internet. Because LGV has established itself in the UK and is present at low levels, surveillance might benefit from supplemented periodic case finding and collection of detailed information of patients to monitor the epidemiology of the disease. Funding The work was produced as part of the PhD study of MR, who was funded by the Osk Huttunen Foundation (Finland) and by the Wellcome Trust (090285/Z/09/Z). HW has received funding for research on LGV from the UK Medical Research Council and support from the National Institute for Health Research Imperial Biomedical Research Centre. PW thanks the Medical Research Council for centre funding.