Control of drug-susceptible TB is largely dependent on a standard 6-month regimen that has been in use for more than three decades. Shortening the course of treatment would be a major improvement for case management and disease control. This study is one of three reported in the New England Journal of Medicine that provide some insight into the factors that are likely to determine the success of new, treatment-shortening regimens. Each trial was designed to test whether the inclusion of a fluoroquinolone in a modified regimen could shorten the duration of treatment for drug-susceptible TB. An editorial notes that in all three studies, observed culture conversion rates at 2 months were consistent with phase II data indicating that fluoroquinolone-containing regimens were likely to be superior. However, this did not reliably predict sterilising activity and risk of relapse: the shortened regimens were not noninferior. The authors suggest that the disconnect between the phase II data that motivated these trials and the phase III results reinforces the idea that small sample sizes limit the utility of short trials in predicting the success of treatment-shortening regimens. They conclude that although the urgency of the medical need may justify additional clinical trials of experimental TB drugs and drug regimens, there must be a considerable increase in investment in fundamental research in order to develop and validate correlates of durable cure. They add that as demonstrated by these studies, understanding of the science underlying positive clinical outcomes remains rudimentary.