Routinely collected data from clinical sites offering voluntary counselling and testing (VCT) for HIV diagnosis have been used to evaluate the impact of public health interventions; however, there has been considerable diversity in strategies. To gain an understanding of the outcome of these evaluations and provide the basis for considering methodological issues, we reviewed published studies. Search criteria were met by 20 papers that described 38 interventions, of which 29 were media-related and nine were policy changes. Most (25 of 38) were based on comparisons between two time periods, before and during the intervention, while 13 used multiple time points, including nine that adopted regression methods. About a third (13 out of 38) of the evaluations monitored HIV positivity rates and a small number investigated impact according to sex (six), age (five) and whether clients were new or repeat (three). For the 29 media-related interventions, there was an average 53% increase in the number of HIV tests performed during the intervention compared with beforehand. For policy change interventions, a 35% increase was found. Routinely collected data from VCT sites can be used to evaluate the impact of public health interventions, but attention to methodological issues will maximize their value for evaluation purposes.