Scleractinian corals increasingly are studied as geochemical archives of modern- and palaeoclimate, but microsampling for geochemical data is complicated by: 1) the microstructural complexity and spatial variability in skeletal growth in different coral genera; and 2) the rapidity and scale of diagenetic alteration that occurs in living coralla. Geochemical sampling techniques now have spatial resolution into the sub-micrometer to tens of micrometers range, and it is hoped that the spatial resolution can be translated to temporal resolution. This study investigated the effects on geochemical analyses imposed by microstructure and diagenesis in different live-collected coral genera representing somewhat different depositional environments. Suites of samples of four reef-building genera (Acropora, Pocillopora, Goniastrea and Porites) were collected from three adjacent environments in intertidal and subtidal positions near the reef edge at Heron Reef, Great Barrier Reef and studied by means of optical and scanning electron microscopy, combined with vibrational and energy dispersive spectroscopy. The first section of this study compares and documents the microstructure of the four coral genera. Each genus was found to have very different three-dimensional arrangements of microstructural elements, and a new general growth model was proposed for Acropora, to take into account differences in the timing of precipitation of trabeculae and thickening deposits. The results highlight the complexity and spatial variability of skeletal growth in different coral genera. Because microstructural patterns vary in different genera, direct observation of microstructural elements and growth lines are necessary to allow geochemical microsamples to be placed into series that represent temporal sequences with known degrees of time averaging. Coral growth rates (i.e., rates of extension) are discussed to determine the range of temporal relationships that exist between closely spaced skeletal microstructural elements. Such data are necessary in order for coral skeletogenesis to be understood and are critical for constraining microsampling strategies aimed at developing true time series geochemical data at very fine spatial and temporal scales. The second part of the study focused on early diagenetic alteration of the corals, which is an equally important concern for geochemical analysis. Early marine diagenesis was documented in the same live-collected samples of the four common reef-building coral genera. Samples show extensive early marine diagenesis where parts of the coralla less than three years old contain abundant macro- and microborings (sponges, algae, cyanobacteria and fungi) and significant amounts of aragonite, high-Mg calcite, low-Mg calcite and brucite [Mg(OH)2] cements. Many of the cements are associated with microendoliths and endobionts that inhabit recently abandoned parts of the skeleton. The cements are problematic for palaeoclimate reconstruction because geochemical proxies used for paleoclimate studies are meant to reflect ambient seawater chemistry and conditions, but the occurrence of brucite and low-Mg calcite demonstrates how far fluid chemistry in microenvironments within the corals has evolved from ambient seawater. Some Porites lobata specimens have had as much as 60% of the most recently deposited skeletal aragonite (i.e., the part of the skeleton that projects into the layer of living polyps) bored and replaced by low-Mg calcite cement. The low-Mg calcite cement has significantly different trace element ratios (Sr/Ca(mmol/mol) = 6.3 ± 1.4; Mg/Ca(mmol/mol) = 12.0 ± 5.1) than the host coral skeletal aragonite (Sr/Ca(mmol/mol) = 9.9 ± 1.3; Mg/Ca(mmol/mol) = 4.5 ± 2.3), thus providing a serious challenge for Sr/Ca or Mg/Ca based sea surface temperature calculations. This study illustrates that many diagenetic changes that can radically alter important geochemical characteristics of coral skeleton occur very early on the sea floor (i.e., while corals are still alive). Documented cements altered trace element inventories (e.g., Sr and Mg), thus, interfering with the use of those elements in palaeotemperature calculations. Hence, significant diagenetic changes that jeopardise palaeoclimate data do not require long-term diagenesis or meteoric exposure. Some of the diagenetic changes (e.g., calcite filled borings) occur at scales that are very difficult to detect short of visual inspection using SEM. Hence, vetting of coral samples with SEM is required before any sample is subjected to geochemical analysis.