USA TODAY research, reported by Julie Snider (July 15, 2004), to predict which golfers would perform successfully in the 2004 British Open, estimated the relative importance of key parts of the game to be: Scrambling 25%, Driving accuracy 20%, Greens in regulation 20%, Putting 20%, and Driving distance 15%. How those percentages were derived was not presented. Nevertheless, identifying the more important elements of the game of golf can contribute to debate and discussion among players, teachers, and coaches, helping to focus on areas in which players need to be strong to be successful. This study sought to develop a statistical profile of the PGA golfer for the 2003 season, and to determine the relative importance of sub-parts of the game to average 18-hole score. Nineteen statistics for 190 players for the full 2003 season were obtained from the PGA: averages for each player for score/round, total driving, driving distance, fairways hit, ball striking, greens in regulation, putts/hole, sand saves, scrambling, bounce back, birdie average, and eight additional statistics reflecting performance on par 3, 4, and 5 holes. Descriptive statistics were computed for all 190 players, for the 50 having the lowest mean score per round (Top Group), and for the 50 having the highest score per round (Bottom Group). T-tests evaluated differences between the Top and Bottom Groups. Multiple regression evaluated relationships between average scores per round and the 18 other PGA statistics. Multi-stage model building developed explanatory/predictive models using the full field of players for 2003, as well as separately for the Top and Bottom Groups. Applying conceptual criteria, the number of statistics used as explanatory variables were purposively restricted for each analysis to maintain acceptable explanatory variable to number of subjects ratios. Mean scores for the top 50 and bottom 50 players were 70.2 and 72.0, respectively, (P < .0005). Of the 18 statistical categories compared between the Top and Bottom Groups, only driving distance was not significant (P=.302). All others were significant (P < .005), with the top 50 having the better performance in each case. Multiple regression models developed, in general, explained 60% - 90% of the variance in average score. Driving accuracy, greens in regulation, putting, and scrambling emerged as significant (P<.05) explanatory variables across almost all models. Drive for show, putt for dough... has anecdotal appeal, but you must perform well in most sub-parts of the game to score well in golf. Keyword(s): athletics/sports, coaching, performance