The purpose of this study was to compare an objective to a subjective metho
d of tooth shade evaluation using the SP78 sphere spectrophotometer and hum
an visual analysis. Twenty extracted fully developed human maxillary teeth
had their L* values (lightness) read by the SP78 on day 1 and then again on
day 14. Using a blind method, five human evaluators then attempted to matc
h shade tabs from a six-tab experimental Vita shade guide to the same teeth
on days 1 and 14. The SP78 L* values read for each tooth were evaluated fo
r a match between those obtained on day 1 and those read on day 14. The res
ults of human evaluation of tooth shade were compared among the evaluators
and then between themselves over time. Finally, the ability of human evalua
tors to match tooth shade was compared with the results obtained with the S
P78. The SP78 reproduced L* readings within the standard error of the machi
ne (less than or equal to 1.0) in 16 of 20 (80%) teeth. In contrast, intere
valuator agreement expressed as majority agreement (3, 4, or 5 of 5 evaluat
ors agreeing) was only 10 of 20 (50%) teeth on day 1 and 13 of 20 (65%) tee
th on day 14. Intraevaluator agreement over the experimental period ranged
from 20 to 60%. The results of this study confirm that human evaluation of
tooth shade is unreliable and that the SP78 sphere spectrophotometer can pr
ovide a more predictable and accurate method of evaluating tooth shade in v
itro. This finding opens up a new avenue of investigation for testing the e
ffectiveness of materials and techniques in bleaching discolored teeth.