This paper examines the relationships between time of transfer from community college to four-year college and three outcomes of attending four-year college: probability of baccalaureate, time to baccalaureate, and credits earned at baccalaureate. Using enrollment data from the State University of New York, I establish that transfer students from community colleges have a longer and more widespread time-to-baccalaureate distribution than do non-transfer (“native”) students, are more likely to have taken time off than natives with the same number of terms completed, and also have fewer credits accumulated than equivalent natives. Multinomial Logit estimations show that “transferring up” after the first, second, and fourth semester of community college attendance is associated with the highest probability of eventual baccalaureate receipt. However, when I add controls for time taken off and credits accumulated, there is a “U-shaped” impact of time spent at a two-year college on the probability of earning a baccalaureate within eight years, with transferring after five semesters now yielding the lowest likelihood of 8-year baccalaureate. Those who transfer “early”—having completed fewer than four community college terms—and do not quickly attrite do not maintain the credit and enrollment statuses of equivalent natives; those who transfer “late”—after six or more semesters at community college—typically earn more credits in a semester than equivalent natives. Without controlling for pre-transfer attendance factors, transferring after four semesters appears to yield the shortest time-to-baccalaureate for those who eventually graduate; when controls for demographics and credits accumulated are added, transferring after exactly one semester yields a significantly higher time-to-baccalaureate for engineering and biological science majors. Transferring up after four or more semesters is associated with extra credits at baccalaureate receipt for students in both math/science majors and majors in large occupational fields, though the impacts are not particularly large in important fields of study.