Selective laser excitation was used to measure the radiative lifetimes of the ZnI triplet states 4s n s 3S 1 (n = 5 -7) and 4s n d3D 3, 3D2 and 3D 1 (n = 4 -6). These states were excited from the metastable 4s4p3P states, which were collisionally populated in an atomic beam. The values are compared with the results of other experimental methods (beam-foil, pulsed electron excitation, Hanle effect) and with theoretical calculations. The corresponding oscillator strengths are discussed with respect to the astrophysical determination of the Zn photospheric abundance.In a recent paper [-1] Biemont and Godefroid have published extensive calculations of oscillator strengths for neutral zinc and its isoelectronic sequence. They have used the multiconfigurational Hartree-Fock (MCHF) programs of Froese-Fischer [-2,3] considering the two outer electrons as a twoelectron system outside a closed core and including pair correlations between these electrons. One reason for this theoretical study was the discrepancy of photospheric and meteoric abundance of zinc. With the new theoretical values of oscillator strengths of the relevant transitions in ZnI this discrepancy is now resolved [4]. It is therefore very interesting to check these theoretical oscillator strengths by new lifetime measurements, because experimental values of lifetimes in Zn I obtained by different experimental methods like the beam-foil method [5,6], the delayed coincidence method with pulsed electron excitation [-7] or the Hanle-effect method [-8, 9, 11] differ considerably. It is well known, that in the case of nonselective excitation methods like the beam-foil method or pulsed electron excitation, cascade effects can reduce the accuracy of the measurements, even if the decay curves are analysed with several decay constants to account for cascade effects. That is why we used selective laser excitation in our measurements. We have concentrated on tripled states starting from the metastable 4s4p3p-states (Fig. 1), which were collisionally excited by electron impact.