ABSTRACT If the photospheres of solar-type stars represent the composition of circumstellar disks from which any planets formed, spectroscopic determinations of stellar elemental abundances offer information on the composition of those planets, including smaller, rocky planets. In particular, the C/O ratio is proposed to be a key determinant of the composition of solids that condense from disk gas and are incorporated into planets. Also, planets may leave chemical signatures on the photospheres of their host stars by sequestering heavy elements, or by being accreted by the stars. The presence, absence, and composition of planets could be revealed by small differences in the relative abundances between stars. I critically examine these scenarios and show that (i) a model of Galactic chemical evolution predicts that the C/O ratio is expected to be close to the solar value and varies little between dwarf stars in the solar neighborhood; (ii) spectroscopic surveys of M dwarf stars limit the occurrence of stars with C/O to ; and (iii) planetesimal chemistry will be controlled by the composition of oxygen-rich dust inherited from the molecular cloud and processed in a dust-rich environment, not a gas with the stellar composition. A second generation of more reduced planetesimals could be produced by re-equilibration of material with dust-depleted gas. Finally, I discuss how minor differences in relative abundances between stars that correlate with condensation temperature can be explained by dust-gas segregation, perhaps in circumstellar disks, rather than planet formation.