The effect of mean gloss level on perceived whiteness was evaluated by magnitude estimation and pair wisecomparison of papers with varying shades and gloss levels. The samples were printed on substrates ofdifferent gloss to target L*a*b* values measured under a light booth illumination having a 5000 K correlatedcolour temperature. Observers were able to rate the whiteness of the samples with large mean glossdifferences in the 5000 K illumination and in the same illumination with an additional UV lamp. The CIEwhiteness equation predicted well the perceived whiteness in both illuminations and the mean gloss level hadno significant effect on perceived whiteness. This means that the CIE whiteness equation can be used tocompare the perceived whiteness of papers having very different mean gloss levels. However, due to thedifferent amount of fluorescent whiteness agents (FWA) in the papers, the perceived whiteness predictionwas only valid when the measurement was performed under the same illumination as for the visualevaluations. Typical indoor illumination with florescent light tubes contains much less UV than the D50 or Cilluminants. The instrumental whiteness differences due to mean gloss and instrument geometry werenegligible compared to the effect of the UV content of the illumination.