Presupernova models and nucleosynthesis in massive stars are reviewd in the context of supernovae. First, presupernova evolutionary models of massive stars toward the onset of collapse from 13 to 70 M(.) stars in the main-sequence stage are presented. It is stressed that silicon (''Si'') shell burning is the key point to determine the final size of the iron core. As the result, we obtain the iron core ranges from 1.2 to 1.9 M(.) for 13-70 M(.) stars. Silicon-rich layers and the boundary between ''Si'' and oxygen-rich layers are neutronized during ''Si'' shell burning. The s-process during core helium burning is investigated in detail for whole massive stars. It is found that the weak component of the s-process is well overproduced compared with the solar system s-nuclei. Next, the results of the explosive nucleosynthesis in massive stars are presented. The explosion energy is assumed to be 1-1.5x10(51) erg. The location of the mass cut is adjusted to explain enough amount of Ni-56 to reproduce the observed light curves of Type Ib, Ic and Type II-L, II-P supernovae. Thus, the mass cut ranges from 1.2 to 1.9 M(.) for our presupernova models. The iron peak elements produced are critically influenced by both the location of the mass cut and the degree of neutronization around there. Then, it is found that the amount of Ni-56 produced ranges from 0.07 to 0.15 M(.). If the neutron-rich elements are observed in a remnant of supernova due to the explosion of a massive star, they would pose severe constraints to the presupernova models and to the scenario of the explosive nucleosynthesis in supernovae. It is suggested that the lower rates of electron captures during ''Si'' burning will save the overproduction of the neutron-rich elements in spite of the way of convective mixing. The abundances ejected by the supernova explosion consist of those produced during the hydrostatic evolution and those by the explosion. For typical abundances (16 less than or equal to A less than or equal to 62) classified as the explosive and hydrostatic burning products, the ratio between the produced and solar system abundances agrees with each other within a factor of 2-3 when we consider the contribution from both Type Ia and Type II supernovae. The p-process during the passage of a shock wave through the oxygen/neon-rich layer co-produces about 60 % of the p-nuclei. It is insisted that the p-process is crucially affected by the seed nuclei produced by the s-process during the hydrostatic evolution of massive stars. Our presupernova models have been obtained using the reaction rate of C-12(alpha,gamma)O-16 published in 1985 which is about 2 times higher than the rate in 1988. Since the presupernova models are influenced by this rate, the ratio of Type Ia/Type II supernovae is also affected. Using the Schwarzschild criterion for convection, our results indicate that the higher rate in 1985 is preferred. However, if convective overshooting is included during core helium burning, a smaller C-12(alpha,gamma)O-16 rate would be suitable. To determine several uncertain nuclear reactions will clarify the confusion between the convection and the reaction rate. The absolute amounts of O-16 and Fe-56 become critical to model the chemical evolution of galaxies and to infer the boundary of stellar mass between a neutron star and a black hole. Thus our results play a crucial role to solve the problems for the origin of the elements, nuclear gamma-ray astronomy, supernova explosion, birth of a neutron star or a black hole, and evolution of galaxies from quasars. Our results will be also very useful to identify the progenitor and predict the observed abundances of supernovae like SN 1987A, SN 1993J and SN 1994I.
