Compared with the impressive progress in understanding signal transduction pathways and mechanisms in mammalian systems, advances in protozoan signalling processes, including cyclic nucleotide metabolism, have been very slow. This is in large part connected to the fact that the components of these pathways are very different in the protozoan parasites, as confirmed by the recently completed genome. For instance, kinetoplastids have no equivalents to the mammalian Class I adenylyl cyclases (ACs) in their genomes nor any of the subunits of the associated G-proteins. The cyclases in kinetoplastid parasites contain a single transmembrane domain, a conserved intracellular catalytic domain and a highly variable extracellular domain - consistent with the expression of multiple receptor-activated cyclases - but no receptor ligands, agonists or antagonists have been identified. Apicomplexan AC and guanylyl cyclase (GC) are even more unusual, potentially being bifunctional, harbouring either a putative ion channel (AC) or a P-type ATPase-like domain (GC) alongside the catalytic region. Phosphodiesterases (PDEs) and cyclic-nucleotide-activated protein kinases are essentially conserved in protozoa, although mostly insensitive to inhibitors of the mammalian proteins. Some of the PDEs have now been validated as promising drug targets. In the following manuscript, we will summarize the existing literature on cAMP and cGMP in protozoa: cyclases, PDEs and cyclic-nucleotide-dependent kinases.