We investigated the effects of diagenetic alteration (dissolution, secondary aragonite precipitation and pore filling) on the distribution of U in live and Holocene coral skeletons. For this, we drilled into large Porites lutea coral-heads growing in the Nature Reserve Reef (NRR), northern Gulf of Aqaba, a site close to the Marine Biology Laboratory, Elat, Israel, and sampled the core material and porewater from the drill-hole. In addition, we sampled Holocene corals and beachrock aragonite cements from a pit opened in a reef buried under the laboratory grounds. We measured the concentration and isotopic composition of U in the coral skeletal aragonite, aragonite cements, coral porewater and open NRR and Gulf of Aqaba waters. Uranium concentration in secondary aragonite filling the skeletal pores is significantly higher than in primary biogenic aragonite (17.3 ± 0.6 compared to 11.9 ± 0.3 nmol · g −1, respectively). This concentration difference reflects the closed system incorporation of uranyl tri-carbonate into biogenic aragonite with a U/Ca bulk distribution coefficient (K D) of unity, versus the open system incorporation into secondary aragonite with K D of 2.4. The implication of this result is that continuous precipitation of secondary aragonite over ∼1000 yr of reef submergence would reduce the coral porosity by 5% and can produce an apparent lowering of the calculated U/Ca - SST by ∼1°C and apparent age rejuvenation effect of 7%, with no measurable effect on the calculated initial U isotopic composition. All modern and some Holocene corals (with and without aragonite cement) from Elat yielded uniform δ 234U = 144 ± 5, similar to the Gulf of Aqaba and modern ocean values. Elevated δ 234U values of ∼180 were measured only in mid-Holocene corals (∼5000 yr) from the buried reef. The values can reflect the interaction of the coral skeleton with 234U-enriched ground-seawater that washes the adjacent granitic basement rocks. We conclude that pore filling by secondary aragonite during reef submergence can produce small but measurable effects on the U/Ca thermometry and the U-Th ages. This emphasizes the critical importance of using pristine corals where the original mineralogy and porosity are preserved in paleooceanographic tracing and dating.