Understanding the mechanisms of transcriptional control by cis-acting regulatory elements is one of the major goals in current molecular biology. Estimates of enhancer numbers in mammalian genomes extend to hundreds of thousands, outnumbering genes by an order of magnitude or more. Yet current high-throughput methods for enhancer identification give only indirect insights into their function, and are unable to demonstrate heterogeneity, redundancy, or co-operativity between regulatory elements. The α-globin locus has been extensively characterized with regard to transcriptional output, chromatin marks and chromosome conformation throughout erythroid maturation, and offers an excellent model for the investigation of enhancer function. Such analysis at this locus has previously been undertaken in heterologous transgenic models, including a 'humanized' mouse bearing the entire human α-globin cluster. Through the investigation of a rare naturally occurring mutation in a patient with thalassaemia, we demonstrate that heterologous transgenic models do not accurately recapitulate normal transcriptional control at this locus. This allows us to revise the prevailing view of human α-globin transcription control, in which a reliance on a single regulatory element (HS-40, or R2) was assumed. An analysis of the transcriptional contribution of the five regulatory elements at this locus (R1-4 and R[m]) has therefore been undertaken, using knock-out mouse models. Assessment of the haematological impact of each regulatory element knock-out demonstrates that none is indispensable for α-globin transcription; however, analysis of double knock-outs proves functional non-equivalence of the elements, with R1 and R2 being critically required for viability. Quantification of nascent transcripts from each model shows a significantly different contribution to α-globin transcription from each of the five putative enhancers, with two providing no transcriptional contribution despite chromatin marks and transcription factor binding profiles suggestive of enhancer activity. R1 and R2 are defined as the minor and major enhancers for α-globin; their deletion in concert also shows an unexpected impact on nascent transcription more widely across chromosome 11, challenging current views on enhancer-promoter specificity. This locus-specific study highlights heterogeneity in regulatory element biology that would be undetectable in current genome-wide studies. The models generated provide a platform for the further investigation of general principles of enhancer function, and will also permit us to examine the regulation of α-globin transcription in non-erythroid tissues.