Quality control has been playing an increasingly important role in the implementation of nucleic acid amplification techniques (NATs) for clinical diagnosis since the introduction of these methods in the early 1990s. Initial multicenter studies involving hepatitis B virus (HBV), hepatitis C virus (HCV), Mycobacterium tuberculosis, and human immunodeficiency virus type 1 (HIV-1) revealed serious problems in specificity (false-positive rates of ca. 40%) and sensitivity, large variations in quantitative results, and a plethora of units (largely not comparable between assays). The problem areas identified included the need for standardized reagents and common units, contamination control mechanisms, inhibition control mechanisms, genotype-independent detection and quantitation, facilitated nucleic acid isolation procedures, clinically relevant dynamic ranges, and internal run controls. Progress made in each of these areas will be discussed. In addition to the above-mentioned problem areas, the value of external quality control of existing and evolving NATs was recognized. To this end, the European Union Quality Control Concerted Action for Nucleic Acid Amplification in Diagnostic Virology was established in May 1998. During its three-and-a-half years of existence, a total of 14 proficiency panels containing 8-13 well-characterized, simulated clinical samples of various viral loads and genotypes were prepared for herpesviruses (herpes simplex virus, human cytomegalovirus), blood-borne viruses (HBV, HCV, HIV-1), enteroviruses, and Chlamydia trachomatis, distributed to up to 20 different countries, and tested by up to 97 different laboratories. The results show dramatic improvement in specificity (false-positive rates <3% for most panels), presumably due to a generally greater expertise of participating laboratories, more frequent use of enzymatic or mechanical contamination control mechanisms, and increased utilization of standardized reagents (commercial kits). However, considerable problems with sensitivity remain (false-negative rates up to 50%), reflecting the high detection limits of some commercial viral load kits still on the market as well as inadequate standardization of quantitation controls between assay systems. In conclusion, although considerable progress has been made, quality control of NATs in clinical diagnosis remains an ongoing challenge.