Macular pigment (MP) is composed of two dietary carotenoids, lutein and zeaxanthin, and a carotenoid generated by the retina, meso-zeaxanthin. There is large intersubject variability in peak optical density, spatial profile, and lateral extent of macular pigment, and it has been suggested that foveal architecture may play a role in this variability. This study is an initial investigation of the relationship between the spatial profile of macular pigment and foveal architecture. Sixty normal subjects were enrolled (one was eventually excluded). The spatial profile of macular pigment optical density (MPOD) was measured by customized heterochromatic flicker photometry (cHFP). High-resolution macular thickness maps were obtained by optical coherence tomography. Four parameters were analyzed: (1) minimum foveal thickness (MFT) at the intersection of six radial scans; (2) central foveal thickness (CFT) averaged over the central 1 mm of the fovea; (3) foveal width identified as the region lacking a nerve fiber layer; and (4) foveal width measured from crest to crest. Lifestyle and vision information were obtained by questionnaire. The mean +/- SD MPOD at 0.25 degrees eccentricity was 0.49 +/- 0.23 and at 0.5 degrees eccentricity, 0.41 +/- 0.21. A first-order decreasing exponential function accounted for most of the variance of the MP profile averaged across subjects (r(2) = 0.99). MPOD measured at 0.25 degrees was unrelated to both measures of foveal thickness for the entire study group (r = 0.03, P = 0.81, and r = -0.08, P = 0.57, respectively). Similarly, MPOD measured at 0.5 degrees was unrelated to foveal thickness in the entire study group (r = 0.12, P = 0.36 and r = -0.05, P = 0.71, respectively). However, when analyzed separately in the nonwhite subjects, the relationship between MPOD at 0.25 degrees and MFT was positive and significant (r = 0.59, P = 0.01), but remained unrelated to CFT (r = 0.20, P = 0.41). Similarly, in the nonwhite subjects, the relationship between MPOD at 0.5 degrees and MFT was positive and significant (r = 0.68, P < 0.01), but again was unrelated to CFT (r = 0.23, P = 0.32). There was no significant relationship between MPOD and either measure of foveal thickness in the white subjects. In the entire study group, there was a positive and significant relationship between foveal width and MPOD averaged across the fovea (r = 0.41, P < 0.01) and between foveal width and MP integrated across the fovea (r = 0.41, P < 0.01). Foveal MP was positively and significantly related to foveal width in the entire study group. This relationship may be determined by the greater length of the cone axons (Henle fibers) in wider foveas. MPOD was unrelated to foveal thickness in the white subjects. However, in the nonwhite subjects there was a positive association between MFT and MPOD at the 0.25 degrees and 0.5 degrees eccentricities, suggesting that other personal characteristics modulate the MPOD-retinal thickness relationship.