Background: In the spring of 2020, the COVID-19 pandemic led to a nationwide unplanned transition to distance learning. For teachers, these school closures resulted in a rapid, substantial transformation of their work. To date, however, almost all of the evidence surrounding the effects of COVID-19 on teachers has been collected retrospectively, after schools had closed. Purpose/Research Questions: Our paper provides, to the best of our knowledge, the first direct evidence of teachers' work prior to and after schools closed as a result of COVID-19. We leverage unique longitudinal data on teachers' daily work activities and affect collected using the Day Reconstruction Method (DRM), which we collected at four time points across the 2019-2020 school year. The DRM asks respondents to provide an exhaustive time budget of their work day, combined with accounts of their positive and negative affect that are tied to specific work activities. This study addresses three questions: 1. How did teachers' time use change during the initial period of the pandemic, in terms of overall number of hours worked and teachers' allocation of time, across an exhaustive set of instructional activities? 2. How did teachers' affect while working, including their positive affect, negative affect, and situational motivation, change during the initial period of the pandemic? 3. How did the relationship between specific teaching activities and affective responses changed during the initial period of the pandemic? Setting and Participants: The data collection took place in two large suburban districts in Connecticut and Rhode Island, as part of an ongoing study on teacher time use and affect. Approximately 250 teachers provided detailed reports of their daily activities at up to four occasions -- October 2019, November 2019, February 2020, and May 2020. All K-12 teachers were invited to participate in the study, and the sample includes teachers in elementary, middle, and high school settings, proportional to the district population of teachers. Intervention/Program/Practice: This descriptive study sought to understand general patterns of time use and affect across and within teachers. We caution against describing this study as a natural experiment, as it is impossible to tease out how school closures and other dynamics related to the pandemic impacted teachers. However, we do have data prior to the pandemic and during its initial months, providing a rich, detailed account of how teachers' experiences changed throughout this time. Data Collection and Analysis: On the DRM, participants report a detailed account of their day, including each of their activities at work and their duration. From this list, the DRM samples five random activities and the participant reports where they were, who, who they were with, and how they were feeling during the activity. The affect items were drawn from the Positive and "Negative Affect Scale" (Watson et al., 1988), a validated tool commonly used in psychological research. Ten of these items can be aggregated to identify a momentary "Positive Affect" score; the other ten can be aggregated into a momentary Negative Affect score. Teachers' reports of time use engaged in specific activities were aggregated to the day level in eight categories: planning, instruction, non-instruction with students, grading/assessment, meetings with teachers, meetings with administrators, meetings with parents, professional development, and other. Analysis began with descriptive accounts of mean reports of time use and affect. Roughly 39% of the variation in time use owed to persistent differences across teachers whereas the remaining 61% owed to variations across days within teachers. For positive affect, about 47% of the variation was attributable to persistent differences among teachers, 11% for days and 41% for activities within a day. For negative affect about 29% was attributable to teachers, 16% for days and 55% for activities within a day. Separate multilevel models were run predicting time use and affect separately. We examined the effect of COVID-19 for each outcome, accounting for teacher variation with days within teachers. Findings/Results: Overall, our emerging findings point to a dramatic re-structuring of teachers' work days following pandemic-related school closures. Perhaps the greatest change is the amount of time teachers were able to spend with students overall. Whereas classroom instruction had accounted for approximately 58% of teachers' time at work prior to school closures, afterward, it accounted for only 18%. Following school closures, teachers spent significantly more time planning, along with grading and assessment activities. In addition, there was a significant increase in the amount of time that teachers spent interacting with others outside of students, including administrators, teachers, and parents. Taken together, these findings suggest a re-orienting of how, and with whom, teachers spent their work days post-COVID. Equally as important, we see significant changes in teachers' experiences of their work. There was an overall decline in teachers' positive affect while working, although negative affect was largely unchanged. But these overall patterns mask changes in how teachers experienced specific work activities after the transition to distance learning. While COVID did not change some aspects of teachers' work (e.g., planning was associated with lower levels of positive affect prior to and after school closings), it appears to have had an effect on others. Most notably, of all of teachers' activities, they appear to be most positive while teaching, and the transition to distance learning appeared to actually strengthen this association. The limited time that teachers spent with students, post-COVID, appeared to be the highlight of their workdays. Conclusions: Given the scope of our longitudinal data on teachers' practice and affect, we believe these findings make a critical and complementary contribution to the literature, describing in detail the changes in teachers' daily work in the initial months of the pandemic. If the pandemic represented a massive exogenous shock to the work of teachers, our study is uniquely poised to shed light into teachers' experiences during a rapid transition. While the changes in time use are somewhat to be expected, the strong link between teachers' positive affect and working directly with students, especially post-pandemic, is surprising and speaks to the value that teachers derive from the "psychic rewards" of teaching (Lortie, 1975).