Performance analysis in complex sports like soccer requires the study of the influence of the interaction between both teams during the game on final performance. The mixed methods approach involves the collection, analysis, and interpretation of qualitative and quantitative data for the same purpose and within the framework of the same study. To build certain observation tools, mixed methods are necessary in order to take advantage of integration between qualitative and quantitative elements. The aim of this study was to develop a new no standard observation tool to analyze soccer offensive performance considering not only the observed team but also some aspects of the opponent behavior, as well as to test its reliability. The process consisted in expert meetings and exploratory observations. Experts carried out several design and re-design steps of the observation tool to its final form which includes two macro-criteria and 31 dimensions. The basic unit of analysis was the 'team possession' and the main aims of study were: (a) technical, tactical and spatial characteristics of the start, the development and the end of the team possession and its offensive performance, (b) the behavior of the observed team just after losing the ball possession and its defensive performance. Inter-observer and intra-observer analysis were carried out and kappa coefficient was calculated to test the observation tool reliability and improve the quality of data. Results indicate that optimal inter and intra-reliability levels obtained in this work are high enough as for suggesting that the observation tool for offensive performance in soccer (REOFUT) could be an adequate tool for analyzing offensive play actions and their performance in soccer.