South African doctors (n = 211) experienced in antiretroviral therapy use were asked via an online questionnaire about the WHO 2013 adult antiretroviral integrated guidelines, as well as clinical and personal issues, in three hypothetical scenarios: directing the Minister of Health, advising a family member requiring therapy amidst unstable antiretroviral supplies, and where doctors themselves were HIV-positive. Doctors (54%) favoured the 500 cells/μl WHO initiation threshold if advising the Minister; a third recommended retaining the 350 cells/μl threshold used at the time of the survey. However, they favoured a higher initiation threshold for their family member. Doctors were 4.9 fold more likely to initiate modern treatment, irrespective of their CD4 cell count, for themselves than for public-sector patients (95%CI odds ratio = 3.33-7.33; P<0.001, although lower if limited to stavudine-containing regimens. Doctors were equally concerned about stavudine-induced lactic acidosis and lipoatrophy. The majority (84%) would use WHO-recommended first-line therapy, with concerns split between tenofovir-induced nephrotoxicity (55%), and efavirenz central nervous system effects (29%). A majority (61%), if HIV-positive, would pay for a pre-initiation resistance test, use influenza-prophylaxis (85%), but not INH-prophylaxis (61%), and treat their cholesterol and blood pressure concerns conventionally (63% and 60%). Over 60% wanted viral loads and creatinine measured six monthly. A third felt CD4 monitoring only necessary if clinically indicated or if virological failure occurred. They would use barrier prevention (83%), but not recommend pre-exposure prophylaxis, if their sexual partner was HIV-negative (68%). A minority would be completely open about their HIV status, but the majority would disclose to their sexual partners, close family and friends. Respondents were overwhelmingly in favour of continued antiretrovirals after breastfeeding. In conclusion, doctors largely supported adult WHO guidelines as public policy, although would initiate treatment at higher CD4 counts for their family and themselves. Resistance to INH-prophylaxis is unexpected and warrants investigation.