Allergic rhino-conjunctivitis (AR) is a globally relevant health disorder characterized by sneezing, rhinorrhea and sleep disturbance. Ragweed (Ambrosia artemisiifolia) is a plant common to North America and an important allergen. Coarse methods of measuring airborne pollen counts are used to predict seasonal allergy symptoms. This research used a longitudinal study design with a novel, model-based raster of predicted pollen counts to test associations with self-reported symptoms of AR collected from patients receiving immunotherapy for pollen allergies at an allergy clinic. Researchers visited a clinic six times over three weeks. Immunotherapy patients were asked to fill out a brief intake survey on allergic and symptomatic profiles, daytime sleepiness, housing quality, and demographics. Participants responded to a daily, emailed survey on sleepiness and asthma symptoms for 21 days. Using the date and location of responses, ragweed pollen counts were extracted from a prognostic, model based raster (25km pixels). Lag associations of pollen counts with sleepiness were tested using a logistic regression model , adjusted for housing and demographic characteristics, in a distributed lag non-linear model (DLNM) framework. 49 people participated in the study. 26 (52%) were female. The mean age was 37.9 years. Asthma/allergy symptoms were not associated with ragweed pollen but sleepiness was highest two days after exposure (Estimate: 0.33 [0.04,0.62]). Subjects traveled widely during the study period. Intense exposures to ragweed pollen may be associated with daytime sleepiness within small exposure windows. Model-based predicted pollen counts could be used to study health impacts of pollen in people with disease severe enough to receive immunotherapy. Daytime sleepiness can affect productivity and injury risk, and pollen season length and allergenicity may be increasing with climate change. Thus our results may have important implications for population health.