Measurements of samples made by two or more spectrocolorimeters, even when made with what the manufacturer certifies as identical instruments, usually do not agree. One source of such disagreements is differences in the physical makeup of the instruments. The optics, physical structure, light sources, spectral analyzers, and detectors of two instruments are never exactly the same. A second source is the interaction of optical characteristics of the sample (e.g., translucency) with the optical configuration of the instrument. In particular, measurement of small areas, such as slightly translucent print control strips, can highlight the problems of measurement disagreement.Measurements of slightly translucent white samples made with two of the commonly used portable instruments, the Gretag SPM series and the X-Rite model 938, often exhibit significant differences. This paper details application to graphic arts measurements of an experimental method of reducing errors caused by interaction of highly translucent plastic samples with instrument geometry. In particular, the work reported here investigates the use this method to modify data taken with a Gretag instrument to make it agree more closely with data taken with an X-Rite instrument.The conclusion of this investigation is that while this method can do a partial correction for differences caused by the interaction of moderately translucent samples with the instruments, the level of uncertainty in measurement data for nearly opaque samples caused by instrument noise, etc. makes its regular use somewhat questionable.