This study was designed to answer two questions: Will the percentage of utility for a group of phonic generalizations in a college-level vocabulary sample be similar to generalizations in a primary-level sample? Will qualitative differences in the types of phonic generalizations be apparent? A 1,495-word sample was selected from George Feinstein's "Programmed College Vocabulary 3600"; Clymer's original list of 45 phonic generalizations was adapted for use in the study and formed the basis for a measure of utility for word pronunciation. The following changes were evaluated for their effects on results: lowering the percentage of utility, prorating the length of the study, and using the separate pronunciation of word parts, in addition to phonic generalizations. It was concluded that phonic generalizations had a lower degree of utility when applied to a college-level vocabulary sample than when applied to a primary-level vocabulary sample. However, generalizations demonstrated a similar degree of utility when applied to a college-level sample and to a sample that was beyond primary level. Tabulations of data and Clymer's list of phonic generalizations are included. (Author/KS)