Background: Substantial evidence demonstrates the harms of pregnancy smoking. Due to the need for very large data sets, less data exist to inform decisions about when quitting should occur, or whether a certain amount of reduction is beneficial. Objectives: Our goal was to examine the effect of timing and amount of pregnancy smoking on low birth weight (LBW) and preterm birth (PTB). Methods: Data for 3 years of registered births in Tennessee were available and included standard birth certificate information. Results: Of 241,616 women, 15.7% smoked at conception, and those who quit as early as the first trimester had significantly increased adjusted odds of LBW (27%) and PTB (14%) compared to nonsmokers, with this effect largely driven by smoking 10+ cigarettes per day. Smoking into the second trimester but quitting prior to the third also predicted increases in LBW and PTB compared to both not smoking at all and smoking only in the first trimester. Smoking to delivery predicted a 2.8-fold increased likelihood of LBW, and a 2.1-fold increased rate of PTB. Any level of smoking after the first trimester substantially increased the odds of poor outcomes compared to both nonsmoking and smoking only in the first trimester. Conclusions: Findings suggest that to avoid LBW or PTB, pregnant smokers should be advised that quitting completely by the end of the first trimester is important, and that continuing to smoke even <5 cigarettes per day after that point substantially increases the potential for of adverse outcomes.