There is a need to explore the relationship, if any, between the probability of a crash and the hours worked by truck drivers. The need arises from the continued adjustment of federal hours of service regulations for truck drivers. This research used data logs from less-than-truckload carrier operations in 2004 to 2005 and in 2010 to estimate the probability of a crash after a certain amount of time spent driving, given no crashes until that time. Driver logs for 7 days before each crash were used and compared with a random sample (two drivers) of drivers who did not crash and were selected from the same company, terminal, and month. This study involved 686 subjects, including 224 crash-involved drivers. Discrete-time survival analysis models indicated a consistent increase in crash odds as driving time increased beyond the fourth hour. Breaks from driving reduced crash odds by as much as 50% compared with situations of drivers with no breaks. Crash odds were lowest when drivers returned to work during the day without an immediately preceding extended recovery period (but with at least minimum required off-duty time). Drivers returning to work immediately after a 34-h recovery period had crash odds 50% to 150% higher than those for drivers without the recovery immediately before a trip. Drivers had the highest crash odds immediately after returning from the extended time off; the effect then diminished with time.