Understanding factors that predict pain experienced by subjects provides a way to better comprehend neural mechanisms involved in pain modulation. Several studies describe predictors for clinical pain, but few have attempted to predict experimental pain ratings using psychological factors and individual differences in sensory thresholds. To investigate predictive factors for experimental thermal pain sensitivity and pain tolerance, twenty-one healthy volunteers were enrolled for the study. Thermal pain thresholds, thermal detection thresholds, intensity, and unpleasantness VAS (Visual Analog Scale) ratings to heat and cold stimuli applied to ventral forearm were assessed. Self-assessment of pain sensitivity, State-Trait Anxiety Inventory (STAI), CES-D (Center for Epidemiological Studies Depression Scale), and PANAS-X (Positive and Negative Affect Schedule – Expanded Form) were completed prior to quantitative sensory testing. Principal component factor analysis was used to reduce a 22 variable predictor set to 5 composite predictors: warm detection threshold, pain positivity (self-assessment of pain sensitivity and positive affect), inter-threshold range (the difference of heat and cold pain thresholds), cold detection threshold, and negative mood (negative affect, STAI, and CES-D). Prediction of pain ratings was then made by multiple linear regression. A model of five composite predictors successfully predicted cold pain sensitivity (r² = 0.614, P < 0.008) and heat pain sensitivity (r² = 0.537, P < 0.027). Individual variable predictors were also used separately. Suprathreshold heat pain stimuli ratings were neither significantly predicted by subjects' self-assessment of pain sensitivity (r² = 0.068, P < 0.252) nor heat pain threshold (r² = 0.058, P < 0.291). Interestingly, cold pain threshold proved to be a good predictor for suprathreshold heat pain stimuli rating (r² = 0.221, P < 0.0317). These results suggest a combination of psychological and quantitative sensory tests may provide powerful multifactorial predictive models for experimental pain with much improvement over individual predictive variables.