Periodization of swim training in the final training phases prior to competition and its effect on performance have been poorly described. We modeled the relationships between the final 11 weeks of training and competition performance in 138 elite sprint, middle-distance, and long-distance swimmers over 20 competitive seasons. Total training load (TTL), strength training (ST), and low- to medium-intensity and high-intensity training variables were monitored. Training loads were scaled as a percentage of the maximal volume measured at each intensity level. Four training periods (meso-cycles) were defined: the taper (weeks 1 to 2 before competition), short-term (weeks 3 to 5), medium-term (weeks 6 to 8), and long-term (weeks 9 to 11). Mixed-effects models were used to analyze the association between training loads in each training meso-cycle and end-of-season major competition performance. For sprinters, a 10% increase between ~20% and 70% of the TTL in medium- and long-term meso-cycles was associated with 0.07 s and 0.20 s faster performance in the 50 m and 100 m events, respectively (p < 0.01). For middle-distance swimmers, a higher TTL in short-, medium-, and long-term training yielded faster competition performance (e.g., a 10% increase in TTL was associated with improvements of 0.1-1.0 s in 200 m events and 0.3-1.6 s in 400 m freestyle, p < 0.01). For sprinters, a 60%-70% maximal ST load 6-8 weeks before competition induced the largest positive effects on performance (p < 0.01). An increase in TTL during the medium- and long-term preparation (6-11 weeks to competition) was associated with improved performance. Periodization plans should be adapted to the specialty of swimmers.