The dmi1/pollux, dmi2/symrk, castor, nup85, nup133, nena and cyclops nodulation signalling mutants are all defective for calcium spiking but retain calcium influx (Shaw & Long, 2003; Miwa et al., 2006b; Groth et al., 2010). However, genes required for calcium influx but not for calcium spiking have not been identified. We measured calcium influx in four mutants of Medicago truncatula, hcl-1, nin-1, bit1-1 and nsp2-2 that retain calcium spiking but lack normal infection threads. The hcl-1 mutation inactivates LYK3, the proposed Nod-factor receptor required for infection thread growth (Smit et al., 2007). The nin-1 transcription-factor mutation causes increased root hair deformation, reduced infection and a failure to form nodules (Marsh et al., 2007). The bit1-1 mutation affects the ERN-1 transcription factor required for nodulation and the initiation of infection threads (Middleton et al., 2007) and the nsp2-2 mutation affecting a GRAS domain transcription factor, causes defects in infection and cortical cell division following inoculation with Sinorhizobium meliloti (Oldroyd & Long, 2003). All four mutants induced calcium influx in at least some of the cells tested (Fig. 3a). Since NIN, ERN and NSP2 are all transcription factors required for infection, it is likely that they would act downstream of the calcium signalling events. As observed previously (Wais et al., 2000; Oldroyd & Long, 2003; Marsh et al., 2007), calcium spiking was normal in these mutants (Fig. 3a). LYK3, a predicted Nod-factor receptor, could have been required for the calcium influx but apparently is not, based on the induction of calcium influx in some cells of the hcl-1 mutant (Fig. 3a). The hcl-1 allele causes a G > E change in the conserved kinase domain (Smit et al., 2007), so possibly some residual function of the protein could be retained allowing calcium influx under the assay conditions. Additionally there seems to be redundancy because, in the M. truncatula hcl mutant, calcium spiking is normal (Wais et al., 2000 and Fig. 3a), whereas in Lotus japonicus, calcium spiking and influx are blocked by mutation of NFR1 (the possible LYK3 orthologue; Miwa et al., 2006b).