Biological rhythms are ubiquitous in eukaryotes, and the best understood of these occur with a period of approximately a day – circadian rhythms. Such rhythms persist even when the organism is placed under constant conditions, with a period that is close, but not exactly equal, to 24 h, and are driven by an endogenous timer – one of the many ‘biological clocks’. In plants, research into circadian rhythms has been driven forward by genetic experiments using Arabidopsis. Higher plant genomes include a particularly large number of genes involved in metabolism, and circadian rhythms may well provide the necessary coordination for the control of these – for example, around the diurnal rhythm of photosynthesis – to suit changing developmental or environmental conditions. The endogenous timer must be flexible enough to support these requirements. Current research supports this notion most strongly for the input pathway, in which multiple photoreceptors have been shown to mediate light input to the clock. Both input and output components are now related to putative circadian oscillator mechanisms by sequence homology or by experimental observation. It appears that the pathways linking some domains of the basic clock model may be very short indeed, or the mechanisms of these domains may overlap. Components of the first plant circadian output pathway to be identified unequivocally will help to determine exactly how many output pathways control the various phases of overt rhythms in plants.