While the subjective measures of hunger and satiety are often used in the context of appetite research, their correlations with food intake seem to vary substantially, ranging from tenuous to moderate. The purpose of the present study was to utilize multilevel mixed models with marginal and conditional correlation coefficients to quantify the relationship between (a) the perception of pre-meal hunger and subsequent intake, (b) the change in hunger across a meal and the amount consumed, (c) the perception of satiety following a meal and the amount consumed at that meal, and (d) the change in satiety across a meal and the amount consumed while accounting for between subject and within subject variability. One hundred fourteen participants were asked to self-report on hunger and satiety before and after lunch and dinner via monopolar numerical rating scales as well as record the weight of each meal via portable scales on three different days. The average consumption per meal was 359 grams (95% CI [328, 390]). Gender had a significant effect on consumption (p = 0.001) as males consumed a mean of 436 grams (95% CI [384, 488]) compared to females (323 grams, 95% CI [287, 359]). Meal also had a significant effect on consumption (p = 0.007) as participants consumed more at dinner (393 grams, 95% CI [359, 426]) compared to lunch (325 grams, 95% CI [291, 359]). When accounting for the random effect of participant ID, which factored in between subject and within subject variability, our multilevel mixed-effects models explained 50-55% of the variability observed in consumption and displayed significant effects on intake for gender (p = 0.0011), meal (p < 0.0001), pre-hunger (p < 0.0001), change in hunger (p < 0.0001), pre-satiety (p = 0.0005), and change in satiety ( p < 0.0001); however, when not accounting for the random effect of participant ID, the models could only account for 4-9% of the variability observed in consumption. Between subject variability contributed to approximately 45% of the random effect of participant ID, whereas within subject variability contributed roughly 55%. This suggests that when only accounting for between subject variability, the predictive power of hunger and satiety on intake in our models would have been somewhat limited. It is likely that our models captured a number of factors affecting within subject variation to some degree. Without being able to measure and disentangle their effects on intake, it is difficult to discern the true effects of hunger and satiety on consumption. As it stands, when accounting for between and within subject variability, hunger and satiety were moderately predictive of food intake in our sample.