There are conflicting data regarding the impact of serum potassium and magnesium levels on susceptibility to ventricular premature complexes (VPCs) in the clinical setting. The associations of serum potassium and magnesium levels with the prevalence of complex or frequent (> 30/hour, multiform or repetitive) VPCs were examined after adjusting for age, sex, smoking, caffeinated coffee consumption, alcohol consumption, and left ventricular mass in Framingham Offspring Study subjects who were free of clinically apparent heart disease. There were 3,327 eligible subjects (mean age 44 years). Complex or frequent VPCs were present in 183 subjects (5.5%). When age-adjusted prevalences of complex or frequent VPCs were compared among quartiles of serum potassium and magnesium using a trend test, lower potassium (p = 0.002) and lower magnesium (p = 0.010) levels were associated with higher prevalence rates of arrhythmia. In logistic regression analyses that included potassium and magnesium simultaneously, potassium (p = 0.0021) and magnesium (p = 0.0311) levels were inversely associated with the occurrence of complex or frequent VPCs after adjustment for age, sex, smoking, coffee and alcohol consumption, diuretic use, and systolic blood pressure. These associations remained significant after accounting for left ventricular mass. A 1 SD decrement in potassium (0.48 mEq/liter) or magnesium (0.16 mEq/liter) level was associated with a 27% (95% confidence interval 6% to 51%) and a 20% (95% confidence interval 3% to 41%) greater odds of complex or frequent VPCs, respectively. Lower levels of serum potassium and magnesium were concurrently associated with higher prevalence rates of ventricular arrhythmias.