In cases of sustained-virological response (SVR or cure) after an ultra-short duration (? 27 days) of direct-acting antiviral (DAA)-based therapy, despite HCV being detected at end of treatment (EOT), have been reported. Established HCV mathematical models that predict the treatment duration required to achieve cure do not take into account the possibility that the infectivity of virus produced during treatment might be reduced. The aim of this study was to develop a new mathematical model that considers the fundamental and critical concept that HCV RNA in serum represents both infectious virus (V<
sub>
i<
/sub>
) and non-infectious virus (V<
sub>
ni<
/sub>
) in order to explain the observation of cure with ultrashort DAA therapy. Established HCV models were compared to the new mathematical model to retrospectively explain cure in 2 patients who achieved cure after 24 or 27 days of paritaprevir, ombitasvir, dasabuvir, ritonavir and ribavirin or sofosbuvir plus ribavirin, respectively. Fitting established models with measured longitudinal HCV viral loads indicated that in both cases, cure would not have been expected without an additional 3?6 weeks of therapy after the actual EOT. In contrast, the new model fits the observed outcome by considering that in addition to blocking V<
sub>
i<
/sub>
and V<
sub>
ni<
/sub>
production (?~0.998), these DAA + ribavirin treatments further enhanced the ratio of V<
sub>
ni<
/sub>
to V<
sub>
i<
/sub>
, thus increasing the log (V<
sub>
ni<
/sub>
/V<
sub>
i<
/sub>
) from 1 at pretreatment to 6 by EOT, which led to <
1 infectious-virus particle in the extracellular body fluid (i.e., cure) prior to EOT. This new model can explain cure after short duration of DAA + ribavirin therapy by suggesting that a minimum 6-fold increase of log (V<
sub>
ni<
/sub>
/V<
sub>
i<
/sub>
) results from drug-induced enhancement of the V<
sub>
ni<
/sub>
/V<
sub>
i<
/sub>
.