The early evolution of metazoans is a major focus of biological attention, but is the historical record revealed in the Cambrian ''explosion'' an accurate reflection of original events? The key questions concern the nature of the earliest animals and when they originated. One widely-mooted suggestion is that planktotrophic larvae, typified by the annelidan trochophore and echinoid pluteus, existed long before the metazoan radiations evident in the Cambrian fossil record. This idea could be consistent for recent evidence of divergence times, based on molecular ''clocks,'' of phyla appearing well before the Cambrian. Now a surprising new discovery of eggs with blastomeres and embryos with well-defined anatomy from the Cambrian (c.530 Myr ago) of China and Siberia promises to extend the arena of debate. In one case a convincing ontogeny can be traced from eggs to adult tube-dwelling cnidarians. In the other example a possible protostome, unhatched and wrapped around the egg, shows segmentation and possibly nascent sclerites. In both, these cases development is direct, i.e., there is no evidence for any planktotrophic larval stage. The implications for our perceptions of both the Cambrian 'explosion' and metazoan phylogeny could be considerable.