Purpose: To determine the frequency and severity of visual function loss in female carriers of X-linked retinitis pigmentosa (XLRP).

Design: Case series.

Participants: Two hundred seventy-six XLRP carriers with cross-sectional data (n = 242) and longitudinal data (n = 34; median follow-up, 16 years; follow-up range, 3-37 years). Half of the carriers were from RPGR- or RP2-genotyped families.

Methods: Retrospective medical records review.

Main outcome measures: Visual acuities, visual field areas, final dark adaptation thresholds, and full-field electroretinography (ERG) responses to 0.5-Hz and 30-Hz flashes.

Results: In genotyped families, 40% of carriers showed a baseline abnormality on at least 1 of 3 psychophysical tests. There was a wide range of function among carriers. For example, 3 of 121 (2%) genotyped carriers were legally blind because of poor visual acuity, some as young as 35 years. Visual fields were less affected than visual acuity. In all carriers, the average ERG amplitude to 30-Hz flashes was approximately 50% of normal, and the average exponential rate of amplitude loss over time was half that of XLRP males (3.7%/year vs. 7.4%/year, respectively). Among obligate carriers with affected fathers, sons, or both, 53 of 55 (96%) had abnormal baseline ERG results. Some carriers who initially had completely normal fundi in both eyes went on to experience moderately decreased vision, although not legal blindness. Among carriers with RPGR mutations, those with mutations in ORF15, compared with those in exons 1-14, had worse final dark adaptation thresholds and lower 0.5-Hz and 30-Hz ERG amplitudes.

Conclusions: Most carriers of XLRP had mildly or moderately reduced visual function but rarely became legally blind. In most cases, obligate carriers could be identified by ERG testing. Carriers of RPGR ORF15 mutations tended to have worse visual function than carriers of RPGR exon 1 through 14 mutations. Because XLRP carrier ERG amplitudes and decay rates over time were on average half of those of affected men, these observations were consistent with the Lyon hypothesis of random X-inactivation.