The first part of this paper deals with the impact of nonsolar and - for late-type, dwarf, and high redshift galaxies - generally subsolar abundances on the interpretation of observational data for starburst galaxies. It points out the differences in colors, luminosities, emission lines, etc. obtained from a model using low metallicity input physics for a starburst on top of the stellar population of a galaxy as compared to an otherwise identical model using solar metallicity input physics only. The second part deals with the chemical evolution during a starburst and contrasts model predictions with observational clues.