This report examines the association between tetrahydrocannabinol (THC) plasma levels and pain response in a secondary analysis of data from a recent diabetic neuropathy study that demonstrated a dose-dependent reduction in spontaneous and elicited pain at specific time points. A randomized, double-blinded, placebo-controlled crossover study was conducted in sixteen patients with painful diabetic peripheral neuropathy. Subjects participated in four sessions, separated by 2 weeks, during each of which they were exposed to one of four conditions: placebo, or 1%, 4%, or 7% THC dose of cannabis. Baseline assessments of spontaneous and evoked pain were performed. Subjects were then administered aerosolized cannabis or placebo and pain intensity and cognitive testing at specific time points for 4 hours. A blood sample was drawn from the left antecubital vein for plasma assay of total THC at 0, 15, 30, 45, 60, 150, and 240 minutes. Associations were made between pain intensity, cognitive impairment and THC plasma levels in this secondary analysis. Results suggested a U-shaped relation whereby pain ratings are greatest at extreme (low and high) levels of THC. The therapeutic window appeared to fall between 16 ng/mL and 31 ng/mL THC plasma level. There was a significant linear effect of THC on only one out of the three cognitive tests. These findings stress the importance of measuring cannabinoid plasma levels when performing future research. Perspective: This analysis correlating plasma THC levels and pain reduction in diabetic neuropathy suggest a therapeutic window. Low and high THC levels had a negative association (no reduction) and THC levels within the window had a positive association (reduction). There was a minor negative linear effect of THC on cognitive function.