Accomplished throwers conserve angular momentum when distal joints of the throwing arm reach peak velocity at a later time than their proximal neighbors. The result is an increase in velocity of the most distal segment--the hand. Past research indicates that skill level varies by the number of joints experiencing distal timing lag (time to peak velocity of a distal joint minus the time to peak velocity of its proximal neighboring joint) and that the amount of lag may vary within skill level across individual performers (Southard, 2002). The purpose of this study was to determine if the amount and variance of joint timing lag differ between and within skill level and age of performer. Eighty participants were divided into four groups according to their age and skill level. Participants were required to throw a ball at 50% maximum velocity. Differences in the timing of three-dimensional joint lag were determined using a 2 x 4 (Age x Level) multivariate analysis of variance (MANOVA). Significant MANOVA was followed by discriminant function analysis, univariate analysis of variance (ANOVA), and analysis of covariance. Coefficient of variance in joint lag was analyzed using separate 2 x 4 ANOVAs. MANOVA results indicated a significant two-way interaction. Discriminant analysis showed that wrist lag was the best discriminating variable for age. Wrist and elbow lag were best for throwing level. ANOVA for wrist lag indicated a decrease in distal timing lag for older and higher level throwers. Elbow lag increased with age and skill level. Coefficient of variance results indicated that joint lag for younger and lower level performers was most variable. It was concluded that such differences in lag should be considered when defining throwing performance. (Contains 4 tables and 3 figures.)