A 10 μm spot of argon laser light was focused onto the outer segments of intact mouse rods loaded with fluo-3, fluo-4 or fluo-5F, to estimate dark, resting free Ca2+ concentration ([Ca2+]i) and changes in [Ca2+]i upon illumination. Dye concentration was adjusted to preserve the normal physiology of the rod, and the laser intensity was selected to minimise bleaching of the fluorescent dye. Wild-type mouse rods illuminated continuously with laser light showed a progressive decrease in fluorescence well fitted by two exponentials with mean time constants of 154 and 540 ms. Rods from transducin α-subunit knock-out (Trα–/–) animals showed no light-dependent decline in fluorescence but exhibited an initial rapid component of fluorescence increase which could be fitted with a single exponential (τ∼1–4 ms). This fluorescence increase was triggered by rhodopsin bleaching, since its amplitude was reduced by pre-exposure to bright bleaching light and its time constant decreased with increasing laser intensity. The rapid component was however unaffected by incorporation of the calcium chelator BAPTA and seemed therefore not to reflect an actual increase in [Ca2+]i. A similar rapid increase in fluorescence was also seen in the rods of wild-type mice just preceding the fall in fluorescence produced by the light-dependent decrease in [Ca2+]i. Dissociation constants were measured in vitro for fluo-3, fluo-4 and fluo-5F with and without 1 mm Mg2+ from 20 to 37 °C. All three dyes showed a strong temperature dependence, with the dissociation constant changing by a factor of 3–4 over this range. Values at 37 °C were used to estimate absolute levels of rod [Ca2+]i. All three dyes gave similar values for [Ca2+]i in wild-type rods of 250 ± 20 nm in darkness and 23 ± 2 nm after exposure to saturating light. There was no significant difference in dark [Ca2+]i between wild-type and Trα–/– animals.