HIV rapid diagnostic tests (RDTs) are now used widely in non-laboratory settings by non-laboratory-trained operators. Quality assurance programmes are essential in ensuring the quality of HIV RDT outcomes. However, there is no cost-effective means of supplying the many operators of RDTs with suitable quality assurance schemes. Therefore, it was examined whether photograph-based RDT results could be used and correctly interpreted in the non-laboratory setting. Further it was investigated if a single training session improved the interpretation skills of RDT operators. The photographs were interpreted, a 10-minute tutorial given and then a second interpretation session was held. It was established that the results could be read with accuracy. The participants (n = 75) with a range of skills interpreted results (>80% concordance with reference results) from a panel of 10 samples (three negative and seven positive) using four RDTs. Differences in accuracy of interpretation before and after the tutorial were marked in some cases. Training was more effective for improving the accurate interpretation of more complex results, e.g. results with faint test lines or for multiple test lines, and especially for improving interpretation skills of inexperienced participants. It was demonstrated that interpretation of RDTs was improved using photographed results allied to a 10-minute training session. It is anticipated that this method could be used for training but also for quality assessment of RDT operators without access to conventional quality assurance or training schemes requiring wet samples.