NC State’s involvement in the NSF-sponsored SUCCEED Coalition has led to a number of changes to the freshman year of the engineering curriculum as reported previously (e.g., ASEE 1999, Porter, et al.). An explicit objective of these changes was to retain in engineering those students who were qualified and interested in engineering, but were leaving engineering for other reasons. While a number of isolated innovations have been studied and have demonstrated positive benefit, this study looks at each freshman cohort from 1987 through 1998 to evaluate changes in retention in engineering during that period. Eleven cohorts were studied; five (1987- 1991) experienced no influence from SUCCEED-sponsored innovations, three (1992-1994) had subsets of the cohort involved in various pilot programs, and four (1995-1998) were more thoroughly affected by SUCCEED-sponsored curriculum changes. Aligned with these cohort groupings, the data indicate three different patterns of attrition. The pre-implementation cohorts are characterized by rapid attrition to a retention of 60-65% by the first semester of the sophomore year, and remaining relatively unchanged beyond that point. The transition cohort data show that the steep rate of attrition of the pre-implementation cohorts was mitigated. Continued attrition through the sophomore year, however, resulted in a transition cohort retention rate that was not significantly different from that of the pre-implementation cohorts. The post-implementation data indicate both a slower rate of attrition and a significantly improved retention rate—with 75% of the 1995 cohort retained in the engineering curriculum after eight semesters and 85% of the 1996 cohort retained after six semesters. If the observed trend in engineering retention continues, NC State might be close to the maximum expected retention, after removing uninterested and unqualified students from the population.