Abstract The most reliable estimate of the performance index of a node, defined as the ratio of flow rates of a component in an exit stream and the feed stream, is obtained by a local mass balance around the node using data pertaining to only those flows that are linked directly to it. It is argued that these performance indices provide useful underpinnings for the global mass balance of multi-nodal mineral processing circuits. Based on this reasoning a new 2-tier method of mass balancing is proposed in which errors are distributed over the reconciled data not only with least possible deviations from measured values but also with minimum perturbations of performance indices of the individual nodes. The method entails estimations of performance indices, followed by their incorporation in a set of mass balance equations that splice together in a consistent manner the exit and entry streams connecting the nodes of the network. In Part 1 of this communication, an algorithm is presented for mass balance around a node and computation of the associated performance indices. For demonstration, two sets of plant data, having multi-level hierarchical flow structures are reconciled. The possible role of the performance index for assessing the quality of the measured data and detailed node-by-node analyses of the circuit performance is briefly discussed.