Spatio-triangulation process, based on a multi-sensor block adjustment, is applied to 40 different VIR and SAR images: Landsat-7 ETM+, panchromatic SPOT-4 HRV, multi-band ASTER, RADARSAT (fine, standard, wide modes) and ERS-1. The images were acquired over Rocky Mountains, Canada from different view/look angles (nadir, across- and in-track) creating various intersection geometries in the overlap areas. Only 1:50,000 paper maps were available for this area. A physical multi-sensor geometric correction model and algorithms developed at the Canada Centre for Remote Sensing were used for the processing. Preliminary results of block formed with all images gave positional errors of around 20-26 m. These errors result from medium-resolution sensors (ERS-1, RADARSAT standard/wide modes, Landsat-7 ETM+), from weak intersection geometry between some images, but these errors also include the 1:50,000 map errors (25-30 m).