To evaluate the Portal color sort test (PCST), a new computer-based test of color vision, by comparing it with a series of clinical tests of color vision in normal and color deficient subjects.

Prospective clinical laboratory study.

Fifty-nine subjects with normal trichromatic vision or with congenital color vision defects underwent a series of color vision tests that included the 15-plate Ishihara test, the D-15 Farnsworth-Munsell test (D-15), the Farnsworth-Munsell 100-Hue test (FM 100-Hue), and the PCST under rigorous standardized conditions, as recommended by the respective manufacturers. The PCST generates a numerical discrimination score comparable to the FM 100-Hue.

To test validity, discrimination scores generated by the PCST were compared with scores on the FM 100-Hue. The Spearman rank correlation between discrimination scores on the FM 100-Hue and the PCST was 0.8 (P < .001). Reliability was assessed by asking patients to retake the PCST at a later sitting. Patients retaking the PCST achieved similar scores on their second sitting as on the first. The correlation in the score between the two tests was 0.7 (95% confidence interval [CI]: 0.4-0.9, P < .001). Median (quartiles) time to complete the PCST was 3.1 minutes. This was faster than the FM 100-Hue time (P < .001), but slower than both the Ishihara and the D-15 (both P < .001).

This study suggests that the PCST, a test of color vision deficiency, can be used effectively and reliably as a tool for screening (comparable to the Ishihara plates and the D-15) and grading (comparable to the FM 100-Hue) color discrimination ability.