Current literature shows that poor cost performance in projects has become routine. Research on cost performance has mostly focused on one of the involved parties, either the client or the contractor. Not many research efforts discuss the cost contingency performance of projects. This research discusses the cost performance and cost contingency of projects in the execution phase from the perspective of a client as well as a contractor. Using a case study approach, 95 projects are investigated: 44 client projects and 51 contractor projects. The results show that depending on the perspective, projects can have cost overruns or cost underruns. Comparing the total realized and estimated costs, projects experienced on average about 16% underrun from a client perspective. From a contractor perspective, projects experienced on average up to 2% overrun. The estimated cost contingency in the client's projects was on average 2.64% more than the required cost contingency. The estimated cost contingency in the contractor's projects was on average 5.41% less than the required cost contingency. These differences are explained by pessimism bias and technical reasons on the client's side. On the contractor side, optimism bias and technical and political reasons play a role, resulting in opportunistic behavior. The findings help practitioners to enhance their cost estimates by avoiding both pessimistic bias and optimism bias behavior - for example, by using historical data from earlier projects. Further investigation into the influence of market conditions on cost estimates is suggested.