I present the discovery of three transiting planets (TrES-2, TrES-3, and TrES-4) of nearby bright stars made with the ten-centimeter telescope Sleuth as part of the Trans-atlantic Exoplanet Survey (TrES). TrES-2 is the first transiting exoplanet detected in the field of view of NASA’s Kepler mission. Of the 20 known transiting exoplanets, TrES-3 has the second shortest period, facilitating the study of orbital decay and atmospheric evaporation. Its visible/infrared brightness makes TrES-3 an ideal target for observations to determine the atmospheric composition. TrES-4 has the largest radius and lowest density of the known transiting planets. These three planets have radii larger than that of Jupiter, and the radius of TrES-4 significantly exceeds predictions from models of hot Jupiters, indicating a possible lack of an energy source in these models. I present the results of Spitzer observations of TrES-2. I reject tidal dissipation of eccentricity as an explanation for the inflated radius, and examine the spectrum for evidence of atmospheric absorption. I have monitored 19 fields each containing 6,000–36,000 stars for evidence of transits. I discuss the rejection of six of my candidate transiting systems from an early field that represent examples of the 67 astrophysical false positives that I encountered in Sleuth data. These six false positives highlight the benefit of a multisite survey such as TrES, and also of comprehensive follow-up of transit candidates. As a further example, I present the candidate GSC 03885-00829 from Sleuth data that was revealed to be a blend of a bright F dwarf and a fainter K-dwarf eclipsing binary. This candidate proved nontrivial to reject, requiring multicolor follow-up photometry to produce evidence of the true binary nature of this candidate. The yield of planets from transit surveys is not yet well constrained or understood. There are numerous factors that affect the predictions such as the amount of correlated photometric noise present in the data. Here I present an analysis of my ability to recover fake transits in TrES data. I examined both the automated transit-search algorithm and my own visual identification process. I find the recovery rate of my visual analysis to be 87% for those transit candidates that had a sufficiently high signal-to-noise ratio to be flagged by my transit-search algorithm and readily identifiable by eye.