To differentiate the effects of high energy phosphates, pH, and [H2PO4-] on skeletal muscle fatigue, intracellular acidosis during handgrip exercise was attenuated by prolonged submaximal exercise. Healthy human subjects (n = 6) performed 5-min bouts of maximal rhythmic handgrip (RHG) before (CONTROL) and after prolonged (60-min) handgrip exercise (ATTEN-EX) designed to attenuate lactic acidosis in active muscle by partially depleting muscle glycogen. Concentrations of free intracellular phosphocreatine ([PCr]), adenosine triphosphate ([ATP]), and orthophosphate ([P(i)]) and pH were measured by 31P nuclear magnetic resonance spectroscopy and used to calculate adenosine diphosphate [ADP], [H2PO4-], and [HPO4(2-)]. Handgrip force output was measured with a dynamometer, and fatigue was determined by loss of maximal contractile force. After ATTEN-EX, the normal exercise-induced muscle acidosis was reduced. At peak CONTROL RHG, pH fell to 6.3 +/- 0.1 (SE) and muscle fatigue was correlated with [PCr] (r = 0.83), [P(i)] (r = 0.82), and [H2PO4-] (r = 0.81); [ADP] was 22.0 +/- 5.7 mumol/kg. At peak RHG after ATTEN-EX, pH was 6.9 +/- 0.1 and [ADP] was 116.1 +/- 18.2 mumol/kg, although [PCr] and [P(i)] were not different from CONTROL RHG (P greater than 0.05). After ATTEN-EX, fatigue correlated most closely with [ADP] (r = 0.84). The data indicate that skeletal muscle fatigue 1) is multifactorial, 2) can occur without decreased pH or increased [H2PO4-], and 3) is correlated with [ADP] after exercise-induced glycogen depletion.