Background Depending on the stroke and distances of the events, swim starts have been estimated to account for 0.8% to 26.1% of the overall race time, with the latter representing the percentage in a 50 m sprint front crawl event (Cossor & Mason, 2001). However, it is still somewhat unclear what are the key physiological characteristics underpinning swim start performance. The primary aim of this study was to develop a multiple regression model to determine key lower body force-time predictors using the squat jump for swim start performance as assessed by time to 5 m and 15 m in national and international level swimmers. A secondary aim was to determine if any differences exist between males and females in jump performance predictors for swim start performance. Methods A total of 38 males (age 21 ± 3.1 years, height 1.83 ± 0.08 m, body mass 76.7 ± 10.2 kg) and 34 females (age 20.1 ± 3.2 years, height 1.73 ± 0.06 m, body mass 64.8 ± 8.4 kg) who had competed at either an elite (n = 31) or national level (n = 41) participated in this study. All tests were performed on the same day, with participants performing three bodyweight squat jumps on a force platform, followed by three swim starts using their main swimming stroke. Swim start performance was quantified via time to 5 m and 15 m using an instrumented starting block. Results Stepwise multiple linear regression with quadratic fitting identified concentric impulse and concentric impulse² as statistically significant predictors for time to 5 m (R² = 0.659) in males. With time to 15 m, concentric impulse, age and concentric impulse² were statistically significant predictors for males (R² = 0.807). A minimum concentric impulse of 200–230 N.s appears required for faster times to 5 m and 15 m, with any additional impulse production not being associated with a reduction in swim start times for most male swimmers. Concentric impulse, Reactive strength index modified and concentric mean power were identified as statistically significant predictors for female swimmers to time to 5 m (R² = 0.689). Variables that were statistically significant predictors of time to 15 m in females were concentric impulse, body mass, concentric rate of power development and Reactive strength index modified (R² = 0.841). Discussion The results of this study highlight the importance of lower body power and strength for swim start performance, although being able to produce greater than 200 or 230 N.s concentric impulse in squat jump did not necessarily increase swim start performance over 5 m and 15 m, respectively. Swimmers who can already generate greater levels of concentric impulse may benefit more from improving their rate of force development and/or technical aspects of the swim start performance. The sex-related differences in key force-time predictors suggest that male and female swimmers may require individualised strength and conditioning programs and regular monitoring of performance.