During flood events emergency measures are used to prevent breaches in flood defences. However, there is still limited insight in their reliability and effectiveness. The objective of this paper is to develop a method to determine the reliability and effectiveness of emergency measures for flood defences. The investigation is focused on measures that prevent breaching of a flood defence in a river system; measures to limit and/or close breaches are beyond the scope. To determine the failure probability of flood defences with emergency measures two assessments are made: 1) the reliability of implantation of emergency measures is determined and 2) the effect of the implemented emergency measures on the reliability of the flood defence. For an emergency measure to function correctly three phases need to be passed successfully: ‘Detection’, where weak spots in the flood defence are detected, ‘Placement’, where emergency measures are built on time, and 'Construction’, which is the successful functioning of the emergency measures. The reliability of ‘Detection’ and ‘Placement’ depends on human and organizational reliability and the feasibility of complete placement of the measures within the available time. The reliability of the ‘Construction’ concerns the structural performance of the emergency measure. For a case study along a part of a river in the Netherlands a failure probability of emergency measures for the piping failure mechanism is estimated to be 1/3 per event. The human / organizational reliability during ‘Detection’ and ‘Placement’ proved to be dominant. When translated to dike ring level the failure probability is reduced with about a factor 2. This is largely explained by the length effect: with increasing amounts of weak spots in a flood defence the contribution of emergency measures will decrease.