The inadequacy of current diagnostics for the detection of low worm burdens in humans means that schistosomiasis mansoni is more widespread than previously acknowledged. With the inception of mass drug treatment programmes aimed at disease elimination and the advent of human vaccine trials, the need for more sensitive diagnostics is evident. In this review, we evaluate the merits and limitations of the principal diagnostic methods, namely detection of eggs in faeces; anti-schistosome antibodies in serum; parasite-derived proteins and glycans in serum or urine; parasite DNA in blood, faeces or urine. Only in the baboon model, where actual worm burden is determined by portal perfusion, have faecal smear and circulating antigen methods been calibrated, and shown to have thresholds of detection of 10–19 worm pairs. There is scope for improvement in all the four methods of detection, e.g. the identification of single targets for host antibodies to improve the specificity of enzyme linked immunosorbent assay. Despite recent advances in the definition of the schistosome secretome, there have been no comprehensive biomarker investigations of parasite products in the urine of infected patients. Certainly, the admirable goal of eliminating schistosomiasis will not be achieved unless individuals with low worm burdens can be diagnosed.