Measurements of the evolution with redshift of the number density of massive galaxy clusters are used to constrain the energy density of massive neutrinos and so the sum of neutrino masses mν . We consider a spatially-flat cosmological model with cosmological constant, cold dark matter, baryonic matter, and massive neutrinos. Accounting for the uncertainties in the measurements of the relevant cosmological parameters we obtain a limit of mν < 2.4 eV (95 % C.L.). Constraints on neutrino masses are of great interest for particle physics as well as for cosmology, and thus attract a lot of scientific attention (for recent reviews see Refs. [1, 2, 3,4]). Current upper limits on the sum of neutrino masses, m ν , from cosmological structure formation data [5,6], cosmic microwave background (CMB) fluctuation data [7,8], or combined CMB + large-scale structure data [9,10,11,12,13], are of order an eV [14] (for various limits see Table 1 of Ref. [4]). The number of neutrino species can be constrained from Big Bang nucleosynthesis or by using CMB and large-scale structure data [11,15].High energy physics experiments also constrain neutrino masses, and have measured the number of light neutrino species with high precision [3]. Direct searches for neutrino mass effects in beta decays yield limits in the region of several eV, but the sum over all neutrino masses is almost unconstrained by beta decay and other experiments, mainly due to the weak limit on the tauneutrino mass. The measurement of neutrino oscillations on the other hand constrains the differences between the squared masses of the neutrino mass eigenstates ∆m 2 . With the justified assumption that neutrino masses are non-negative and for mass splittings ∆m , we obtain m ν > 0.04 eV if the solar mass splitting is between the highest and second highest mass eigenstates ( m ν > 0.07 eV if the atmospheric mass splitting is between the two highest states). Results from the LSND collaboration yield a larger lower limit on m ν , and must be considered if confirmed by the MiniBooNE experiment [16], which is currently taking data.In this paper we use the dependence of galaxy cluster number density evolution on the massive neu- * Electronic address: tinatin@phys.ksu.edu † Electronic address: evt@phys.ksu.edu ‡ Electronic address: arna@lukash.asc.rssi.ru § Electronic address: ratra@phys.ksu.edu trino energy density parameter Ω ν to set a limit on m ν . We consider the standard spatially-flat ΛCDM Friedmann-Lemaître-Robertson-Walker cosmological spacetime model with baryons, cold dark matter (CDM), massive neutrinos, and a non-zero cosmological constant Λ (for a recent review see Ref.[17]). To compute the cluster number density as a function of redshift z we use the Press-Schechter approach [18,19] as modified by Sheth and Tormen (ST) [20].1 This approach makes use of the mass function N (M > M 0 ) of clusters (cluster number density as a function of cluster mass M greater than a fiducial mass M 0 ), which depends on cosmological model parameters [21,22,23,24]. In particular,...