We therefore reviewed the medical records of clients attending MSHC who had tested HIV-negative and who subsequently tested HIV-positive at MSHC within 1 year of their last negative test. We reviewed all new HIV-positive cases over a 3-year period from 1January 2004. We extracted epidemiological data and clinical notes relating to risk, interventions and discussions held during consultations. Quantitative data was analysed using SPSS software (SPSS Inc, Chicago, IL, USA). Annotated records were examined and transcribed by two researchers (SP, RC), independently. A process of qualitative thematic analysis then occurred in order to generate common themes. All 20 clients who fulfilled the inclusion criteria were men who have sex with men (MSM) and had a mean age of 32 years. At their last HIV-negative consultation before testing HIVpositive, 16 reported inconsistent condom use for anal sex among 23median male partners (range 1 to 200) over the previous 12 months. Over half (13) clients had at least one documented sexually transmissible infection (chlamydia/ gonococcal/herpes simplex virus) diagnosed at the Centre in the year before HIV positive diagnosis. In the medical records of the HIV-negative consultations, qualitative themes generated relating to their risk of acquiring HIV included drug and alcohol use (n=1), depression (n=6), serodiscordant relationships (n=6) and history of sexual abuse (n=2). Post HIV diagnosis, the medical records indicated the same themes at higher rates and, in particular, drug and alcohol use during high-risk events (n=9). Depression was slightly higher (n=7), as was a history of sexual abuse (n=3). Despite reporting considerable levels of risk for HIV transmission at the last HIV-negative consultation, only seven records documented an offer of counselling and in only five of these cases were the clients referred for counselling or other services to assess risk. Two records indicated that referral to these services had been declined. We found that while these clients often reported unsafe sex, the full context of the sexual encounter was often not documented in the medical record at the time of the negative test but was documented later. Of particular note only one client had drug or alcohol use documented initially, but after the HIV diagnosis, nine had this documented. While we acknowledge medical records may not fully reflect the extent of discussions held during a consultation, these findings indicate that documentation needs to be improved in medical records and also that a more proactive process of referral for counselling should be implemented in the Centre. We are, therefore, implementing a quality improvement process to ensure enhanced documentation of contributing factors that predispose to unsafe sex and for the offer of referral to counselling services and the outcome of that offer. The recommendation is that HIV-negative MSM are asked questions relating to the context of high-risk events and that the offer of appropriate services be made if these co-contributors are acknowledged during sexual history taking. Wepropose that seroconversion to HIV among clients seen at a sexual health service could be used as an important quality marker that should be regularly audited. Given the current rises in HIV seen in Australia, fine-tuning services for MSM is one way of reducing future HIV cases. 1