The purpose of this study was to evaluate the feasibility of using push-ups to predict one-repetition maximum (1 RM) concentric bench press performance. College males (n = 106 enrolled in a basic fitness course were evaluated for a 1 RM bench press with free weights and for maximum one-minute push-up performance. Push-ups were adjusted for body height (PU[middle dot]cm), body mass (PU[middle dot]kg), lean body mass (PU[middle dot]kg LBM) and height and body mass (PU[middle dot]cm[middle dot]kg[middle dot]100-1). Push-ups were only moderately related (r = 0.47) to bench press. The correlations between adjusted push-ups and bench press were r = 0.49 (PU[middle dot]cm), r = 0.71(PU[middle dot]kg), r = 0.65 (PU[middle dot]kg LBM) and r = 0.71 (PU[middle dot]cm[middle dot]kg[middle dot]100-1). Cross-validation (n = 53) of the prediction equation developed from the original sample (PU[middle dot]kg) and the equation proposed by Dean et al. were moderately successful (r = 0.70 for both). However, the standard errors of estimate were +/- 15.7 kg for each equation, although the Dean et al. equation significantly overestimated bench press performance. Because of the large error noted when predicting 1 RM bench press, push-ups are not an accurate reflection of upper body strength in young males.
(C) 1991 National Strength and Conditioning Association