Slug test results are usually treated with scepticism based on the idea that they are fundamentally unreliable. Based on decades of experience, this paper proposes that many tests yield erroneous values of hydraulic conductivity for straightforward practical reasons. The main cause of large errors is leakage from the interval under test both during drilling as well as post-completion. Other causes of error are misguided field procedures, unfortunate design stratagems, and misidentified test behaviour. The paper, based on real examples from various geological environments, shows how to identify problems after the event as well as how to avoid them. The major innovation suggested here is the use of the rate of change of water level, often termed the ‘pressure derivative’, versus elevation (rather than time) as a way of identifying likely leakage. It is concluded that much of the poor procedure arises from ambiguities and omissions in the UK code of practice. The paper suggests a new set of practical guidelines to limit or measure error based on current technology and an integrated approach to test analysis. It is proposed that all future monitoring boreholes are tested with an approach designed to assess potential leakage pathways.