Spark chambers have been used in conjunction with photographic emulsion to investigate the accuracy and reliability in the location of primary cosmic rays in emulsion exposed at balloon elevations. The location of these primaries in the emulsion stack has been determined using measurements made on spark chamber photographs. A sample of 118 primaries withZ>1 which did not interact in the emulsion target has been used to calibrate the positions of the emulsions with respect to the spark chambers. The standard deviations of coordinate location errors in the emulsion plane and perpendicular to the emulsion plane were both 0.2 mm after calibration. The standard deviation of errors in the projected angles in the emulsions for these primaries was 0.2°. The mean and standard deviation of errors in the dip angles for these same events were 0.1 and 0.2°, respectively. A total of 191 out of 192Z>1 primaries, and 24 out of 39Z=1 target interacting primaries have been located in the emulsion without significant ambiguity. All the 63Z=1 noninteracting primaries with energies greater than 50 GeV have also been located in the emulsions without any ambiguities. Scanning efficiencies for theZ>1 primaries and theZ=1 noninteracting primaries were both about 100%. The scanning efficiency for theZ=1 target interacting primaries was approximately 43%. The scanning time required for the location ofZ>1 andZ=1 primaries which did not interact in the emulsion target was 2–3 h. Corresponding scanning times for the location ofZ>1 andZ=1 target interacting primaries were on the order of 6–8 h and two to four days, respectively.