Adoptive T cell therapy (ACT) against cancer utilises cytotoxic T lymphocytes (CTL) specific for tumour associated antigens to target and eradicate cancerous cells. While highly successful in some cancer patients, in others, the therapy fails due to poor CTL effector function and inactivation. We aimed to investigate what determines the success or failure of ACT. By deploying a murine B cell lymphoma model, we were able to show that in-vitro activated, adoptively transferred CTL can eradicate a small number of tumour cells expressing their cognate antigen. However, with increasing numbers of antigen+ tumour cells at the time of ACT, a population of adoptively transferred CTL expanded, which showed reduced tumour killing ability. Cytotoxicity and effector function of this CTL population decreased with higher numbers of antigen+ lymphoma cells, while its expression of inactivation associated molecules increased. When ACT was performed after the number of antigen+ tumour cells reached a certain threshold, all the transferred CTL rapidly showed characteristics of this population and totally lost their tumour killing ability, leading to failure of ACT. Expansion of inactive CTL and loss of effector function was not induced by long-term persistence of antigen but was observed as early as 24 h after the adoptive transfer. Hence, our results describe a rapid inactivation mechanism of adoptively transferred CTL induced by encounter with a high number of antigen+ tumour-target cells. We hypothesise that in patients, the frequency of adoptively transferred CTL with characteristics of the inactivated CTL population shortly after ACT, could be utilised as an early predictor of therapy outcome.